Home

User manual - Institute for Transport Studies

1. where gives the version number of the software Inside this main DRACULA screen are three child windows Animation Text report and Simulation Speed At the start of a simulation the animation window lies on top of the text window By clicking anywhere on any of the other two windows you can bring it to the front The child windows are restricted within the parent window DACULA MARS 1 0 On top of the parent window are four menu bars which provide user interfaces to the simulation program The details of the menu options are described in Section 3 7 2 and Appendix C The animation window displays the simulated network To get a closer view of any part of the network use the left hand button on the mouse to select the area of interest then click on menu option View Box to zoom in The text report window is used to display information about the simulated network and messages regarding the simulation process at the end of the simulation At the start of the simulation the screen is blank DRACULA Manual v2 4 21 Section 3 Executing the Programs MM Text report NETWORK SUPPLY INFORMATION NUMBER SIMULATION JUNCTIONS NUMBER SIMULATION LINKS HUMBER SIMULATION LAHES NUMBER SIGHAL STAGES HUMBER DETECTOR NUMBER BUS LANES HUMBER BUS STOPS THE SIMULATION JUNCTIONS ARE DIVIDED INTO EXTERNAL NODES TYPE amp PRIORITY JUNCTIONS TYPE 1 TRAFFIC SIGHALS TYPE 3 DUMMY H
2. OR input the network full path name e g C DRACWIN test OTLEY NET into the text box _ _ DRACULA WINDOWS CONTROL Programs Folder Module Run Command Line Run Demo EITS E DRACULA DATA PRE PROCESSOR eE ChexPDRACNET NET ITS Step 3 Start the pre process by clicking Run button Alternatively the above steps can be carried out with one command from main menu Command Line Run Prompt or DOS Prompt for example DRACULA Manual v2 4 17 Section 3 Executing the Programs D DRACULA DRACPREP netname DRACULA WINDOWS CONTROL Programs Folder Module Run Command Line Run Demos Tools Help BITS BE COMMAND LINE Please Input Command Line C idractexeldracprepl chworkshop Sex LDRACNET DRACULA Manual v2 4 18 Section 3 Executing the Programs 3 4 Running the Simulation Step 1 Load the network from main menu Module Run DRACSIM Step 2 The dialogue DRACULA MICRO SIMULATION pops up i a al M DRACULA MICRO SIMULATION Load Network ITS Step 3 Click Load Network button to load network OR input the network full path name e g C DRACWIN test OTLEY SUP into the text box Step 4 Start the simulation by clicking Run button Alternatively the above steps can be carried out with one command from main menu Command Line Run Prompt or DOS Prompt for example D DRACULA DRACSIM netname DRACULA Manual v2 4 19 Section 3 Executing
3. LTP 60 XYUNIT 1 FILTRP ottrip1 FILBUS otbus amp END 4 2 2 2 Simulation Node Data Most of the SATURN simulation node data are used directly by DRACULA in the same way as they are used in SATURN The following node data are interpreted into DRACULA data NAME DESCRIPTION JTYPE Type 9 junction for giveway to the right traffic control See Section 4 2 4 JCIR Time to circle roundabout See Section 4 2 2 6 RSAT Roundabout capacity See Section 4 2 2 6 The node data not used by DRACULA is NUC Number of time units per cycle 4 2 2 3 Simulation Link and Turn Data The SATURN simulation link and turn data not used by DRACULA are STACK Link stack capacity QSTAR Marker for speed flow curves LSAT Turn saturation flow See N B 1 and the whole section on Link speed flow data N B DRACULA Manual v2 4 45 Section 4 Basic Highway Traffic Network and Demand 1 Junction or turn saturation flows are determined by car following rules They are outcome of micro simulation models The types of SATURN turn priority parker TPM used by DRACULA are G a give way turn at priority junction X Opposed right turn from a major road M Merge F A permanent green filter BLANK No opposing flow In addition the following TPMs which are specific to DRACULA can also be specified in the SATURN format networks see Section 4 2 1 5 for details S For motorway slip roads Z A zipped turn a
4. C and S to indicate a clear turn to the near and far side exit respectively 4 2 1 5 The vehicle characteristics are drawn from truncated normal distributions with the values of mean variance minimum and maximum of the distributions specified by the users 6 1 3 The default simulation break out time is now set as three times of the main demand period 6 1 1 TECHNICAL Provide warning messages in TXP file if a point along a curved link falls inside the junction 10 1 5 When a curved link is specified in GIS file more than once only the first specification is used and a warning message is written to TXP file 10 1 5 Two menu options to display points along a curved link and their A and B node numbers are provided in the DRACSIM window 3 8 2 Examples of network and other coding and practical tips are highlighted with graphical symbols The author wishes to acknowledge the contribution of Yaron Hollander who created the symbols for examples and practical tips DRACULA Manual v2 4 140 APPENDIXES e Option to do a simulation run for a pre defined time period 6 1 1 e Improved display of stopline marks along curved links and on approaches to roundabouts e Improved display of bus lane and flared approached on curved links e Allow space in direction and file names e Improved display of bus layby they are now drawn offside of the main carriageway e Check on route connectivity
5. Repeat for every vehicle Repeat for every junction Fig 1 1 The simulation loop The network is represented by zones nodes links and lanes A zone is the source or sink of traffic where they enter or leave the network A node is either external connecting to a zone or an intersection connecting to other nodes inside the network In DRACULA there is no restriction on the number of roads connecting an intersection DRACULA Manual v2 4 3 Section 1 Introduction A link is a directional roadway between two nodes and consists of one or more lanes A link is specified by its upstream and downstream nodes cruise speed number of lanes turns permitted to other outbound links from the downstream node and restricted access such as reserved lanes For each permitted turn the lane s in the link that can use this turn are specified and a marker describing its priority over opposing flows is given In the model traffic moves in lanes A lane can be reserved for a particular type s of vehicles only for example a reserved bus lane The reservation is specified by its start and end position on the lane and optionally by a start and end time of its operation See 4 2 Vehicles are individually represented each has a set of individual characteristics including vehicle type car bus guided bus taxi heavy goods vehicle vehicle length desired minimum distance headway normal and maximum acceleration normal and max
6. no rt P instead of no grt P in generate veh pool e Allow SATURN style comment lines lines beginning with a in the GIS inputs routines changed are read gis in t5load c 25 08 2006 e Allow automatic search for detector inputs from filename DET e Separate bus based outputs in out file cf SATURN output format e Error corrected on results of passenger delay outputs in PSN and PAS files e Apply user specific stop line position to all node types as specified via ADD file e Allow off line setting of default colours via parameters such as COL FAC COL LANE etc for the colour of the clock face lane marking etc This provides facility to set preferred colours for all runs DRACULA Manual v2 4 2 as oppose to set them on line during each simulation run The new parameter names are listed in Appendix B DRACULA Manual v2 4 3 TABLE OF CONTENTS 1 1 1 1 2 2 ZA 22 3 1 3 2 3 3 3 4 3 6 Sf 3 8 4 1 4 2 INTRODUCTION ioscan a aaaea 1 THE DRACULA CONCEPT orure E A AE l OVERVIEW OF DRACULA TRAFFIC SIMULATION MODEL 2 SYSTEM REQUIREMENTS AND ARCHITECTURE 7 COMPUTER SYSTEM REQUIREMENTS cccccssccesccssccssceescesceusceesces 7 MODEL ARCHITECTURE AND DATA REQUIREMENTS 7 Dod DRACPREP em a e a E A A nua meyers 9 22a DRACSIM oo a a ean ae ecg 10 Zoe CNES ore asics Ea SD Maric ela anaes A 11 EXECUTING THE PROGRAMS ccccccccecececeeeeeeeer
7. 2 SYSTEM REQUIREMENTS AND ARCHITECTURE 2 1 Computer System Requirements DRACULA is written in C C and compiled with Salford C C and Salford ClearWin for running on PCs under 32 bit Windows systems Windows XP 2000 NT and earlier Windows systems A_ typical specification for a system to run DRACULA would include Table 2 1 System requirements Requirements Hardware 64MB RAM 0 5GB hard disk Windows XP or earlier systems Software library Salflibc dll Screen setting True colour 32 bit 2 2 Model Architecture and Data Requirements There are two main programs in DRACULA MARS 1 a pre processor called DRACPREP which processes the input data and creates the required environment for the microsimulation and 2 a traffic microsimulation program called DRACSIM The data required for DRACULA MARS includes the travel demand data in terms of routes used flow on each route and optionally drivers preferred departure times the network topology traffic controls and simulation control parameters Figure 2 1 shows the architecture of DRACULA MARS and depicts the relationship between the two programs with the input data required and the output data files produced Details of the input and output files are explained in 2 2 1 and 2 2 2 2 2 3 explains file name conventions DRACULA Manual v2 4 J Section 2 System Requirement and Model Architecture Discretionary Input Data Public Transport bus Mand
8. Had it be the other way round then the appropriate region of the bitmap would have been selected and suitably expanded Thus a very useful property of the bitmap displays is that they move with the network window File Display View Simulation ME Animation gresosssseseossssosy AANA SEE SERISEIS RE vo Yosossiy o ma to tay Manual v2 1 Micros cj EE DRACULA MARS 2 0 DRACULA Manual v2 4 122 Example REFERENCES REFERENCES Bonsall P Liu R and Young W 2005 Modelling safety related driving behaviour impact of parameter values Transportation Research 39A 425 444 Clark D J Blythe P T Thorpe N and Bourke A 1993 The ADEPT Project 3 Congestion metering the Cambridge trail Traffic Engineering amp Control April 1993 Clark S D May A D amp Montgomery F O 1996 Priotity management for urban artetials transferability of techniques implementation in NEMIS ITS Technical Note 387 Institute for Transport Studies University of Leeds Department of Transport 1991 New car fuel consumption the official figures December 1991 UK Dtp Ferreira L J A 1982 Car fuel consumption in urban traffic the results of a survey in Leeds using instrumented vehicles ITS Working Paper 162 Institute for Transport Studies University of Leeds Liu R 2000 Simulation of plan selection traffic signal control Report prepared for York City Council Liu R 2005 T
9. Menu Full Network enables the whole simulated network to be displayed in the animation window Full background ditto for background map Window Option to control viewing changes to network only background only or both Box enable the user to get a closer view of any part of the network Use the left hand button on the mouse to select the area of interest then click on menu option Box to zoom in Options Zoom Pan Left Right Up Down enable the user to move to any interested parts of the simulated network The last five options Scale2Pt BMP Drag BMP Set BMP X Y Save BMP X Y and Fit whole BMP to screen enable scaling and positioning of bitmap background map 3 8 2 4 Simulation Menu Display vehicle by provides options to display the colour of vehicles by type junction turning route or randomly Animation by options to display the animation by individual vehicles or aggregated link delay DRACULA Manual v2 4 25 Section 4 Basic Highway Traffic Network and Demand DRACULA Manual v2 4 26 Section 4 Basic Highway Traffic Network and Demand 4 BASIC HIGHWAY TRAFFIC NETWORK AND DEMAND The imput data required to run DRACULA MARS are divided into fundamental data and advanced data The former is essential to run the model while the advanced data allows optional model features to be activated This section provides an introductory discussion on the structure and the fundamental input data files that are
10. PhD thesis Institute for Transport Studies University of Leeds Oldridge B 1990 Electronic road pricing an answer to traffic congestion Proc Information Technology and Traffic Management HMSO London QUARTET Deliverable No 2 1992 Assessment of current tools for environment assessment in QUARTET DRIVE II Project V2018 QUARTET September 1992 Watling D P DRACULA 1 0 user guide to the day to day model ITS Technical Note 369 Institute for Transport Studies University of Leeds 1995 Van Vliet D 1982 SATURN a modern assignment model Traffic Engineering and Control 23 12 578 581 Van Vliet D And Hall M 2002 SATURN 10 2 User Manual Institute for Transport Studies University of Leeds and W S Atkins January 2002 DRACULA Manual v2 4 124 APPENDIXES APPENDIX A Control Parameters for Data Pre Processing DRACPREP incall Charge band in meters in the congestion pricing scheme Coded in as integer value Section 8 5 CHRGSEG Segment in metres in the pricing scheme Coded in as integer value Section 8 5 CIRC RADIUS 4 Minimum radius of roundabout central island metres Section 4 2 2 3 CIRC SPEED 27 Roundabout circulating speed km hr See Section 4 2 2 3 FIJPMIN Minimum flow veh hr below which a route is not selected for DRACULA simulation See definition in Appendix B GAP_MIN DE e Di GAP TSTART 600 GAP TEND Ola A GONZO 1 0 Demand factor for the whole trip matrix
11. The data required to model such features are termed GIS data in line with the terminology used in SATURN The GIS features currently available are enclosed polygons polygon lines icons church BR station etc text and curved links The data to be stored in text file GIS follows exactly the same format to that in SATURN see SATURN Manual Appendix Z In fact the SATURN users can use P1X to generate the GIS features In versions prior to Version 2 1 released in October 2003 the SATURN format GIS files have to be modified slightly before DRACSIM can load them For example the corners of a polygon poly line or a curved link have to be given specifically by the user for DRACSIM The newer versions do not require such information The newer versions can take any GIS file created for earlier versions of DRACSIM programs The following set out the format for specifying GIS features The format is the same as those used for SATURN Those in italic are not used in DRACSIM 10 1 1 Enclosed Polygons Record 1 1 Enclosed Polygons Cols 1 5 11111 Record 1 2 1 Start of a new polygon Cols 1 10 X co ordinate of the first corner Cols 11 20 Y co ordinate of the first corner Cols 21 25 Pen colour in range 1 to 16 Cols 26 30 A non zero 1f the area is to be filled Col 31 35 Name of zone Col 39 50 An alphanumeric title for the area 9 Record 1 2 2 Co ordinates of the polygon s next up to 4 corners Cols 1 10 X co
12. a the area covered by the full network and b the full area covered by the bmp file The full network area can be worked out by the programs from the coordinates of the nodes coded in the network data file The area covered by the bitmap can be defined by the coordinates of its 4 corners Thus for every bmp file used by DRACSIM say picture bmp it is necessary to set up a further very small file named picture xyb which specifies the 4 corners of picture using the same coordinate system as that used by the network The xyb files consist of a single record on the same line containing 4 real values in the following order DRACULA Manual v2 4 121 Section 11 Links with DRACULA Day to Day Model a XMIN the east west co ordinate of the lower left hand corner b XMAX ditto for the upper right corner c YMIN the north south co ordinate of the lower left hand corner d YMAX ditto for the upper right corner The xyb file may be most conveniently set up the user assuming that the information is known in advance through knowing the source of the image Alternatively if a bitmap is input into DRACSIM without a corresponding xyb file being located the full bitmap is drawn with its XMIN YMIN corner to the bottom left hand corner of the screen The example below shows an aerial photograph drawn underneath a network plot In this example the network window displayed is larger than the area covered by the bitmap
13. reserved lane description To code in bus service routes bus stops and reserved bus lanes from scratch follow the format described in the next three sub sections and entered via text file net BUS In addition the following parameters can be set in PAR file to specify bus related variables Table 5 2 Parameters of bus operation QBUS F Set T to allow loading from BUS file PFASTBRD Percentage passengers holding seasonal bus tickets 4 0 Ticket purchasing and boarding time per passenger sec person 1 e constant a in Eq 5 1 Boarding time second person for reason ticket holders 1 e constant a in Eq 5 1 Door opening and closing time for buses sec 1 e the constant b in Eq 5 1 DRACULA Manual v2 4 71 Section 5 Modelling Public Transport 5 2 1 Bus service data Data in this section must be preceded by a single line containing string amp BUS SERVICE and terminated by a single line containing 99999 DATA NAME DESCRIPTION tkk RECORD 1 SERVICE DESCRIPTION BUSNO Name of bus route alpha numeric string of up to five characters 2 TRROUTE One of the following characters entered in the space immediately after BUSNO T if the route is two way R if the route is defined in reverse order otherwise leave blank See Note 1 3 VTYPE Type of buses see Section 6 1 2 4 TSTART Time the first service starts minutes from the beginning of the simulation 5 FREQ Frequency of the
14. 1 3 Vehicle Characteristics There are nine parameters used in DRACULA to describe the physical and behavioural characteristics of each vehicle driver unit They are 1 length metre 2 minimum safety distance metre 3 reaction time second 4 normal acceleration m s2 5 maximum acceleration m s2 6 normal deceleration m s2 7 maximum deceleration m s2 8 desired speed factor 9 gap acceptance factor The users can specify change the distributions of the vehicle parameters via a text file VEH TAB The format of the file together with the default characteristics for the default types of vehicles are listed below The file starts with an identifier amp VEH PARAM and ends with amp END For each type of vehicle there are four rows of nine records representing the nine vehicle characteristics The first row for each vehicle type contains the mean values of its nine parameters The second row contains the coefficients of variation the third contains the minimum allowed values of each parameter and the fourth contains the maximum As an example the following file lists the default values veh tab file amp VEH PARAM CAR 4 5 1 0 1 0 1 5 2 0 2 5 5 0 1 00 1 0 0 1 01 0 0 01 0 1 dado 0 1 0 10 Brad Seo GF sO Papi Dol Lets Oe Das Ds Seo ube el Dad cons Sale 65 2300 Lab BUS Pas bRO Ar Pad tab Teb Ze t00 Das 0 1 0 1 0 0 0 1 0 1 0 1 0 1 0 10 0 1 5 0 0 8 1 0 0 8 0 8 1 0 1 0 0 50 0 2 TOO eLege eb 20 20 420
15. 4 2 1 Start of a new text Cols 1 10 X co ordinate of the centre of text Cols 11 20 Y co ordinate of the centre of text DRACULA Manual v2 4 120 Section 11 Links with DRACULA Day to Day Model Cols 21 25 Pen colour in range 1 to 16 Cols 26 30 Character height in millimetres on the screen Cols 33 60 Text Record 4 2 2 as 3 2 1 for the next text etc Record 4 3 End of text Cols 1 5 99999 10 1 5 Co ordinates of Curved Links Record 7 1 Curved links Cols 1 5 eee Record 7 2 1 Link Identification Cols 1 10 Upstream node of link Cols 11 20 Downstream node of link Record 7 2 2 Co ordinates of the intermediate points Cols 1 10 X co ordinate of the next 1 corner Cols 11 20 Y co ordinate of the next 1 corner Cols 21 30 X co ordinate of the next 2 corner Cols 31 40 Y co ordinate of the next 2 corner etc Record 7 3 End of curved links Cols 1 5 99999 10 2 Bitmap Background Another way to enhance presentation of the study area is to display a background map of the area such as an OS map or an aerial photograph underneath the network plot Currently within DRACULA such background map must be of bmp format as opposed to e g pcx gif etc formats The other graphical formats may be converted into a bmp format by making use of standard software such as Paint In order for DRACSIM to draw a bitmap background within the windowed area covered by a network plot it 1s necessary to know
16. 67 CHE MODEL eoan a e a San 67 Jd AMOU ON Ese a a ra 67 5 1 2 Modelling Public Transport Servie 67 5 1 3 Modelling Bus Stops and Bus Dwell Time 68 5 1 4 Modelling Reserved Lan S roce n e E 70 DATA FILES orana e a isa Era a ca 71 I2 PUSS CAO s E E YZ Ja BU SION AGI wacicte RD RA END RN DR RC ANDRES E RE RR 4 323 Reserved bus lane dal sanar asai ad Usada o ar 75 AN EXAMPLE esses E E 76 GUIDEWAY FOR GUIDED BUSES cccccccceeeeeeeeeeeessnenaaeeeeeeeeeeeees 76 TIME DEPENDANT PASSENGER DEMAND ccecceeceeccsceeseescesees 76 SIMULATION SETTINGS wissiisssecicsssscesssrcessiiinciasccavecteeus 79 GLOBAL SETTINGS a a a aa 19 OJA SMUAIN CLIOGS di 5 ita Rania alii Td ER eects 79 OM VE ELY S ienei tah ada dO RSA a och edad daN 81 6 1 3 Vehicle Characteristics cccccccccccccccccsssssesseseeesesseeeeceseeeeeenaaaas 82 6 1 4 Traffic Composition and Demand Factor ii 83 OAD CarFolowngModeaS saias asse irao ais das asda 84 6 1 6 Gap Acceptance MOdelS iiccccccccccccsssccccccccccccccssessseccseesaaassssesees 84 Ol ELaneChangino Models ccrvsinciivccenszactasterissaieeeteaiGeaderiaeleads 86 6 1 8 Reaction to Traffic Control at Intersections 86 LOCAL SETTINGS ss ssa aa cased eet toes 90 6 2 1 Junction or Link Specific Gap Acceptance Parameters 90 022 Sop line LOCA sis aaa isa a densa al da 91 6 2 3 Reaction zone to JUNCTION control eeeeereereeemeeeees 91 6 2 4 Roundabout SPecifiCAtiOn ccccccccccccccc
17. Ab Te Ab GBUS Tag dns lt b Lodo a6 dad ed 100 Das 0 1 0 1 0 0 0 1 0 1 0 1 0 1 0 10 0 1 5 0 0 8 1 0 0 8 0 8 1 0 1 0 0 50 0 2 10 0 1 2 1 0 2 0 2 0 4 0 4 0 1 50 1 0 TAXI 4 5 1 0 1 0 1 5 2 0 2 5 5 0 1 00 1 0 0 1 041 0 0 01 0 1 0 1 0 1 0 10 Quad S20 Be Leo Lob dO 140 dQ 0450 Oss Sea demo The Bae roel Sud S 2200 2x0 LGV Pas che chet dado eG dS Zu 0 80 440 0 1 0 1 0 0 0 1 0 1 0 1 0 1 0 10 0 1 Deh OO aos CO Be Lob Ers USO OA Det Ger Mo SAO waa A AO GeO ge HGV dade A0 Fs bg ive Lad ud O80 s0 0 1 0 1 0 0 0 1 0 1 0 1 0 1 0 10 0 1 e Ore do Og Oe ds dad OaD Ds 10 0 1 2 1 0 2 0 2 0 4 0 4 0 1 50 2 0 Qh mi Z e J DRACULA Manual v2 4 82 Example Section 6 Simulation Settings 6 1 4 Traffic Composition and Demand Factor When the origin destination O D trip matrix and route assignment for each user class or vehicle type are known they can be loaded onto the simulation according to the description given in Section 4 3 This section describes how mixed traffic can be represented with knowledge of only an average trip matrix and assignment In practice a single trip matrix is often collected combining several types of traffic If one knows the average distribution of user classes vehicle types among the matrix e g how many different types of vehicles there are in the network and on average what proportions of the trip matrix are made up by these vehicle types the following method can be use
18. DRACULA emission and fuel consumption models takes into account explicitly vehicles four different driving modes acceleration deceleration cruising and idling and calculates emission by pollutant and fuel consumption for each individual vehicle based on their instantaneous speed and acceleration The emission factors used are taken from the QUARTET Deliverables No 2 QUARTET 1992 Emission rates for three pollutants are available CO NOx and un burnt HydroCarbon HC emission For vehicles cruising at a constant speed the emission factors grams second are assumed to be a function of speed as shown in the table below Table 7 1 Emission factors for driving at constant speed Pollutant speedikpi co oos 0091 0 13 0 129 009 011 0177 NO 00008 00006 00077 0 062 0 0082 0 00807 0 0058 The emission factors for the cruising mode were given as discrete speed points at every 10kph A linear interpretation is used to get emission factors at speeds in between any of the two listed in the table In other driving modes idling accelerating or decelerating constant emission factors are assumed Table 7 2 Table 7 2 Emission factors for idling accelerating and decelerating sid dling Accelerating Decelerating CO 0 377 0 072 NOx 0 0008 0 005 HC 0 0067 0 0067 The fuel consumption factors are taken from Ferreira 1982 and DOT 1991 It has assumed that fuel consumption factors are constant f
19. FWIDTH should be less than 1 0 However sum of the FWIDTHs of the two way links should be greater or equal to 1 0 iv A vehicle approaching a narrow bridge needs not only look at the vehicles already on the bridge but also those vehicles approaching the bridge on the other side The SIGHT sets out the distance downstream of the link approaching the bridge from the other end by which the subject vehicle would look out Example B o Example 1 Example 2 amp BRIDGE amp BRIDGE 103 104 63 0 5 20 105 104 48 104 103 48 104 1 SR ZZ O wa e 99999 29999 The above Example 1 is used together with the link coding presented in Fig 4 3 to describe that the bridge has a combined width of a normal lane width A give way sign is put up on approach from link 102 103 Fig 4 3 Hence the VTYPE 63 coding here where traffic from link 102 103 will have to give way to all types of traffic They would give way not only to all types of traffic already on link 103 104 but also to all types of traffic 20 metres downstream on link approaching node 104 from the other direction e g link 105 104 Figs 4 1 and 4 3 Even though the give way sign is only put up on node 103 side of the bridge for traffic from link 102 103 vehicles approaching the bridge from link 105 104 should also give way to any vehicle travelling on link 103 104 This is indicated by the VTYPE 63 for link 104 103 DRACULA Manual v2 4 57 Section 4 Basic Highway
20. If all or parts of a bus route follows a straight line more or less 1t is not necessary to define every node along the line but only the nodes at the corners or at the ends The NNODES should be the number of corners coded instead of the actual number of nodes en route C All bus stops en route should be coded Note in SATURN networks public services are coded as routes under card 6 Use this information to help code up the public transport services in the following format If any of the links on the bus routes are really part of a link in the real network as with the reserved lanes de code the bus links to lanes Bus routes coded in SATURN networks can be converted to DRACULA data Following the procedure below If there are bus routes to be extracted from existing SATURN network go to step 1 else go to step 4 Step 1 Step 2 Step 3 Step 4 Step 5 Set parameter QSATNET T QBUS F in net PAR file Run DRACPREP On completion you should have a net BUS file created which has three records beginning with amp BUS SERVICE amp BUS STOP amp BUS LANE Each record ends with a string 99999 There should only be data entries under the first record go to Step 4 for further data editing Set parameter QBUS T in net PAR file Edit text file net BUS following the instructions below to add more or build from scratch bus routes bus stops bus lanes DRACULA Manual v2 4 73 Section 5 Modelling Public Transpor
21. This input detector file indicates that there are four detectors in the network identified by number 5 6 201 and 202 detector 5 lies 50 meters upstream from the stopline of the kerb lane in link 1011 1012 and it is for detection of all types of vehicles detector 6 lies 150 metres downstream from the entry of link 1123 1345 and is placed on the kerb lane of the link It detects buses only and triggers a demand dependant signal control at junction 1345 with maximum recall length of 200 seconds Detector 201 and 202 make a pair of co ordinated detectors for a plan selection signal control at junction 727 the first triggers and the latter cancels the plan selection The alternative signal plan for junction 727 is included in the second section of the data file DRACULA Manual v2 4 101 Section 8 Advanced Simulation Applications 8 2 Responsive Signal Controls There are four types of traffic responsive signal controls modelled in DRACULA they are all responsive controls to selective vehicles e g they can be responsive only to selected types of vehicles such as buses Section 9 7 describes a fifth control type combined pedestrian and traffic responsive signals The four traffic control algorithms are a extension of green signals in order for example to allow a bus pass the stop line without delay b early termination of a red signal in order to reduce bus delays c demand dependent signals which calls an
22. Traffic Network and Demand 4 3 Highway Travel Demand For the current release of DRACULA MARS travel demand is an exogenous input At the most detailed level the demand is modelled at the individual level individual trip maker the route they take on the day preferred departure time and their individual driving behaviour Section 4 3 2 Such information can be generated by the DRACULA day to day route choice model Section 11 or any other microscopic route choice models At a more aggregated level travel demand is represented in terms of average flows vehicles hr over assigned routes and time periods The latter representation of travel demand described in Section 4 3 1 is a more practical one given the current level of data available i e OD trip matrix rather than individual trip makers and can be readily linked with other assignment packages such as SATURN see Section 4 3 3 Different from SATURN and other assignment models traffic flow in DRACULA is represented in terms of number of trips per hour or number of vehicles per hour as opposed to pcus per hour So when converting SATURN flows into DRACULA flows conversion from pcu to trip is required see Section 4 3 3 4 3 1 Average Travel Demand Though greater effort has been made to link DRACULA directly with SATURN assignments it is also possible to link DRACULA with other assignment packages with relatively minimal effort required from the user The users need simpl
23. acceptable gap Section 6 1 6 Waiting time sec before taking on MIN GAP Section 6 1 6 Flow factor Section 6 1 4 Parameters as used in SATURN to control the display of GIS features n 1 2 7 is for GIS polygon polyline icons text name curved link and node coordinate respectively Currently SATURN GIS names and node coordinates are not represented in DRACULA Option IFGIS 5 is always set to True e g the display of curved links are always ON if they are given in GIS file FWARM 0 5 Start level of traffic flow during warm up period as a proportion of the peak demand Section 6 1 1 DRACULA Manual v2 4 128 APPENDIXES O Gi e e NCOOL CYCLE 3 Number of cooling off periods after a signal extension or recall Section 8 4 NSEED 800 Random number seed for all random events other than the generation of random arrival headways Gf QARRIV T Section 4 The number varies automatically with day counter as NSEED NSEED 100 NDAY Section 11 NEED2 4321 Random number seed for generating vehicle arrival headways Section 4 It varies with day counter if QDEMAND T as NSEED2 NSEED2 100 NDAY Section 11 PBUS Proportion of total demand in number of vehicles hr made up by buses i e PBUS 10 10 of total demand are buses Section 6 1 4 PFASTBRD Proportion of total passengers hold seasonal bus tickets PHGV Proportion of total number of trips are made by HGVs Section 6 1 4 PISA Percentage vehicles fleet unde
24. and Junction Layout The coded link length is defined in both DRACULA and SATURN as the distance between the centre of the upstream junction and the centre of the downstream junction along the path of the road The path may be a straight line between the upstream and downstream nodes or it could be curved making the length of the path longer than crow fly distance between the upstream and downstream nodes However unlike SATURN which doe not model the physical size of a junction except for Node Graphics display DRACULA explicitly models the physical layout of intersections and estimates the size of an intersection from the number of lanes of the links and a fixed lane width therefore the effective or inter stop line length of a link in DRACULA is the link length taking away the distance between the centre of the upstream junction DRACULA Manual v2 4 40 Example a Section 4 Basic Highway Traffic Network and Demand to the entry of the link and the distance between the centre of the downstream junction to the stopline of the link It can happen that a negative length is calculated in the above method This could be due to an un realistically short link length coded in SATURN which was used to connect two very close intersections or due to an incorrect interpretation of junction size see Section 4 2 2 3 When a negative link length is detected a warning message is given in the pre processing report file TXP
25. and NRBL for the roundabout concerned via roundabout specific input in ADD file see Section 6 2 4 2 3 Signalised Roundabout Model The roundabout models described in the basic DRACULA networks Section 4 2 1 and in SATURN networks Section 4 2 2 3 represent those operating with give way entries The model described in this section can represent roundabouts which are fully or partially operated by traffic signal controls at the entries In fact the logic developed is generic which allows both give way and signal controlled roundabouts to be represented More detailed representation this roundabout model allows information about the number of lanes lane markings on the roundabout angles of entry exit approaches and traffic control on approaches to be specified In addition this model can represent different characteristics for different sections of a roundabout For example you can model a roundabout partially operates as two lanes and partially operates as three lanes You can also model a partially signalised roundabout some arms operate as priority approaches and others signal controlled For these reasons the new roundabout model is sometimes referred to as a partially signalised roundabout model This model can be used in conjunction with both the basic DRACULA and SATURN format networks This section of data should be entered in VET file after the main network descriptions described in Sections 4 2 1 and 4 2 2 have been cod
26. and all time periods Section 4 3 1 2 GONZO t 1 0 Matrix factor for time period t Section 4 3 1 3 LEFTDR T F if driving on the right e g European driving Section 4 1 MUC v l Vehicle type of user class v Section 4 3 3 2 NMUC Number of user classes Section 4 3 2 Random number seed Section 4 sn BE of time periods tee 4 3 1 2 4 3 1 3 4 3 3 3 ee Per car unit used to convert flow given in pcu or pcu hr to number of vehicles or vehicles hr Default 1 pcu 1 vehicle of type CAR See Sections 4 3 3 1 Section 4 3 3 1 PED COVI Coefficient of variance of the start time of a pedestrian route Section 8 7 DRACULA Manual v2 4 125 APPENDIXES Coefficient of variance of the duration of a PED COV2 pedestrian route Section 8 7 Coefficient of variance of the pedestrian flow Section 8 7 T if to load bus service bus stops and bus lanes from BUS file Section 5 2 1 PED COV3 QBUS QCROWFLY T if to use crow fly distance as link length Section 4 2 2 2 QDETECTOR T if to load detector data from det file Section 8 2 QEXPERT If T the level of print out in TXP file generally is such that only an expert would fully appreciate it QPASSQ F T if to load SATURN multiple time period assignment Section 4 3 3 3 QRBLANE If T the number of lanes on a roundabout is limited to at least equal to the maximum lanes on entry arms F if the circulating lanes are determined from SATURN roundab
27. and bus PSN dwell times Detector records of passing vehicle TXD speed and type Trajectories for selected vehicles 2 2 3 File Names Unless specified otherwise via the method described below all input data files for each network should be stored in the same directory The outputs files from DRACPREP and DRACSIM will be in the same directory Both the file name and the directory name should be a string of alpha numeric characters without a white space in between For example a name should not be named as OTLEY BASE nor should it be stored in directory c Program Files The extensions of all the input and output files used m DRACULA As listed in Tables 2 3 2 5 are fixed For example a file used to specify an input network description has to have an extension of NET and the output pollution measures can only be found in a file with extension POL DRACULA Manual v2 4 11 Section 2 System Requirement and Model Architecture By default all the input data files for the same network should have the same name and it is the name of the NET file For example a network called OTLEY will have its parameter control file named OTLEY PAR trip file OTLEY TRP etc However different names can be used for different input files for the same network This can be done via parameters m the PAR file see Section 4 1 for the format of PAR file Table 2 6 lists the parameters used PARAMETER and the extension of their correspon
28. at which charges are levied may not coincide with those at which major delays occur This raises the question of how to model drivers reactions to charges in terms of their perception of network travel conditions For this reason only a route based individual vehicle simulation model may fully represent the impacts of the most technologically advanced road user charging systems At the time of writing an updated version is being made Potential users may wish to contact the author for a test version DRACULA Manual v2 4 107 Section 8 Advanced Simulation Applications 8 5 Modelling Overtaking on Two Lane Rural Roads The links are classified into two types e links with double white lines where vehicles can not move over to the other side of the road or e links without double white lines where overtaking from the other side of the road 1s permitted By default all links are the first type unless specified otherwise A network can have a mixture of both types of links This section describes the logics behind DRACULA models of overtaking on rural two lane roads without double white lines The model is largely based on the earlier work conducted in Australia and implemented in model TRARR The model first examines drivers overtaking desires If a vehicle falls within a catching up region defined by a distance headway threshold DRUB and relative speed threshold VRUB its speed 1s severely constrained by the slow mov
29. extends to bus routes DRACULA Manual v2 4 141
30. flow time secs Free flow speed kph Distance o a Power x 100 Capacity WIC ratio per cent SPATULA net Delays at jen secs Cruise time secs Cruise Speed kph Total time Gis secs Net speed Gs kph Demand flow CO kilos NOx kilos hydrocarbons Fer cent green split TAS qued vehs mid jen REKS DRACULA Manual v2 4 113 Section 9 Model Outputs and Evaluation In P1X use Display Option Choice of link annotation By List you will get to a window Link Data Set shown below on the right Among those listed the measures between Delay at jcn secs and Hydrocarbons are results from the DRACULA simulation Others are SATURN inputs outputs 9 3 Generalised Cost DRACULA generalised cost C is defined as C PPM xT PPK x D 9 1 where PPM is pence per minute PPK pence per km T total vehicle minutes D total vehicle km DRACULA Manual v2 4 114 Section 9 Model Outputs and Evaluation 9 4 Performance vs Supply Measures DRACULA makes clear distinction between the performance of a network and costs associated with a given demand the supply costs and produce summary statistics for each type of measures The performance of a network or a single link can be measured in terms of flow performed and time performed in a defined period They are engineering description of the performance of the link or network at a given point in time or over a given time period and can be used to estima
31. integer or real parameters In the case of a real parameter value the number can be input as either real value or integer numbers Some of the parameters are used in the pre processing stage others affect the behaviour of the simulation The list of user definable parameters is long many of the variables and their use are complicated The full list of the parameters used in DRACPREP and DRACSIM are given in Appendix A and B respectively with pointers to appropriate sections A subset of parameters given in Tables 4 1 and 4 2 for DRACPREP and DRACSIM respectively is suggested as reasonable starting points where explicit choices need to be made DRACULA Manual v2 4 21 Section 4 Basic Highway Traffic Network and Demand Table 4 1 Parameters defining the options of basic input to DRACPREP 800 NSEED Random number seed An integer number preferably of three or more digits to trigger a pseudo random number generator Used both in DRACPREP and DRACSIM QSATNET T T if to take SATURN format simulation network F to use the free format network coding as described in Section 4 2 2 QSATPIG F F if to take routes and route flows from input file TRP of format defined in Section 4 3 1 Else if T take them from LPG files 4 3 3 QBUS F T if to load bus service bus stops and bus lanes from BUS file Section 5 2 1 LEFTDR F if driving on the right e g European driving Factor converting one unit of node zone coordinates to me
32. is higher by a pre defined factor Section 8 6 describes another special type of lane changing overtaking on a single carriage way using the road space in opposite direction When a vehicle wishes to change lane it looks for a target lane The target lane is generally determined by the lane changing requirement except in the case of overtaking which is only permitted from the nearside to the offside Once it has chosen a target lane it examines the lead and lag gaps in its target lane in a way which is analogous to a gap acceptance model DRACULA allow drivers to anticipate the need for a change of lane It uses variables XAPPRO JNCT and TAPPRO JNCT see Section 6 1 and 6 2 to determine the distance and time to lane changing target point such as a bus stop beginning of an incident or the stop line of a junction in case of types 1 b or c lane changing DRACULA Manual v2 4 86 Section 6 Simulation Settings DRACULA also allows for the willingness to create gaps by kind hearted drivers 1n the target lane Section 6 1 6 and variation in the gap depending on the urgency of the desire to change lanes the closer to the target point the smaller the gaps they would be willing to accept For overtaking DRACULA uses the following two control parameters if the current speed is below F2 OVERTAKE times its desired speed and the speed gain in target lane would be over FI OVERTAKE times its current speed an overtaking de
33. movements both permits its next junction turning movement and leads it onto the lane in presence of a with the bus stop Failing to do so the bus would first get into the next link bus stop then look for gaps to move to the lane with the bus stop Entering in a link with bus stop s a bus would always aim to travel in the lane where the bus stop lies or adjacent lane in case of layby This may involve changing to the lane with the bus stop The bus would only look to get into a lane that permits its next junction turning after it has passed the last bus stop on the link A bus only stops at its own bus stops as listed in its service data A bus in a layby does not affect the traffic on the road In the model a bus which stopped at a layby is removed from the stream of traffic on the road Once it has picked up all the waiting passengers the bus will try to re merge into the traffic stream according to gap acceptance rules The passenger volume at each bus stop is represented by an average hourly flow A which is converted to a uniform rate of arrival The bus dwell time T2 at a bus stop is related to the number of passengers Passenger flow N waiting at the bus stop in the following way and bus dwell time T a l p N apN b 5 1 where a 1s the time 1t takes for one passenger to get on a bus including paying to get a ticket az the time it takes for a seasonal ticket holder to get on the bus p the proportion of passengers usin
34. node A whilst stream a of an S turn enters the motorway acceleration lane at node C D C B A DRACULA Manual v2 4 39 Section 4 Basic Highway Traffic Network and Demand ii C A C turn is used to describe a turning movement to the right or far side for driving on the left which also has a clear exit This is the same as the C TPMod turning modifier used in SATURN Section 4 2 2 3 The general rule for a C turn is that 1 it has its own exclusive exit lane 2 it does not need to give way to traffic streams sharing the same exit however 3 it does need to give way to priority crossing traffic In the following example turning movement a has a clear exit It gives way to conflicting traffic c on the far side but not to traffic b from near side which turns to the same exit link Give way to c on far side not to b on near side Exit to far side lane in exit link iv Z A Z turn works as a G turn at a priority junction in free flow situation During extreme congestion at the priority junction the Z turn and its higher priority counter part will take in turns to enter the junction and merge into the exit link The congestion threshold where such a transition occurs is modelled by a speed threshold if the average speed on the major road is below the threshold the rule changes from G to Z The speed threshold is specified via a global parameter VZTPM Appendix B still under development 4 2 1 6 Link Length
35. of time steps Maximum 10 steps TSTEP 1 o Duration minutes for time period t Duration minutes for time period t 1 2 3 10 QPASSQ F Leave it as F is to assume fixed assignment over NSTEP time period GONZO GONZO t Matrix factor for the whole time period Matrix factor for time period t DRACULA Manual v2 4 61 Section 4 Basic Highway Traffic Network and Demand 4 3 2 Individual Trips By default the DRACULA MARS traffic simulation requires individual drivers route and departure time choice for the day When exogenous input on such detailed demand description is not available the program DRACPREP can still generate a list of individual drivers whose routes are generated based on the aggregated demand provided hence VEH 1s referred to as an output from DRACPREP in Section 2 2 One advantage to feeding a pre defined list of vehicles to DRACSIM as opposed to generating vehicles within DRACSIM is to enable consistency in the driver vehicle characteristics simulated between among different scenario tests To do multiple simulation runs on the same scenario see Section 9 5 however it is advisable not to run DRACSIM with the same list of vehicles To enable DRACPREP to generate net VEH file set in net PAR file the parameter QVEHPOOL T The format of the VEH file contains two records RECORD TYPE 1 Total number of vehicles to travel NVEH entered on the first line of the file RECORD TYPE 2 R
36. opportunistic pedestrians who may cross the road during red pedestrian signal if appropriate gaps available the difference between the two depends on their patience A number of pedestrian actuated signal policies are implemented Pedestrian stage length may vary with number of waiting pedestrians A pedestrian stage may be extended for late crossing or vice verse be terminated early if all pedestrians have crossed the road Similarly a pedestrian stage may be brought forward in excess demand or vice verse be cancelled in absence of pedestrian demand The model is implemented within DRACSIM During the implementation of pedestrian simulation a vehicle actuated signal policy is also represented where a vehicle green stage may be extended at the detection of approaching vehicle s by downstream detectors Interactions between this vehicle actuated and the pedestrian actuated signal policies are dealt with in the model Further tests are to be conducted before its release Prospect users may want to contact the author for a test version DRACULA Manual v2 4 110 Section 9 Model Outputs and Evaluation 9 MODEL OUTPUTS AND EVALUATION 9 1 Summary of Output Files The default outputs from a DRACULA simulation are the link route and network based traffic condition measures and pollution measures at a regular time period and at the end of simulation The output time period is user specified and can be as small as one second The DRACULA
37. ordinate of the next 1 corner Cols 11 20 Y co ordinate of the next 1 corner Cols 21 30 X co ordinate of the next 2 corner Cols 31 40 Y co ordinate of the next 2 corner etc Record 1 3 End of polygons Cols 1 5 99999 DRACULA Manual v2 4 119 Section 11 Links with DRACULA Day to Day Model 10 1 2 Polylines Record 2 1 Polylines Cols 1 5 22222 Record 2 2 1 Start of a new polyline Cols 1 10 X co ordinate of the first corner Cols 11 20 Y co ordinate of the first corner Cols 21 25 Pen colour in range to 16 Cols 26 30 Line width in millimetres on the screen 9 Record 2 2 2 Co ordinates of the polygon s next up to 4 corners Cols 1 10 X co ordinate of the next 1 corner Cols 11 20 Y co ordinate of the next 1 corner Cols 21 30 X co ordinate of the next 2 corner Cols 31 40 Y co ordinate of the next 2 corner etc Record 2 3 End of polylines Cols 1 5 99999 10 1 3 Icons Record 3 1 Icons Cols 1 5 33333 Record 3 2 1 Start of a new icon Cols 1 10 X co ordinate Cols 11 20 Y co ordinate Cols 21 25 Pen colour in range 1 to 16 Cols 26 30 Icon height in millimetres on the screen Cols 31 35 Icon number representing 1 for a monopoly style house 2 fora BR symbol 3 fora car park 4 for a church 5 for a hospital Record 3 2 2 as 3 2 1 for the next icon etc Record 3 3 End of icons Cols 1 5 99999 10 1 4 Text Record 4 1 Text Cols 1 5 44444 Record
38. overtaking links 8 5 and ISA speeds 8 3 DRACULA Manual v2 4 135 APPENDIXES e New SATPIG with SATURN 10 3 outputs to DRACULA required TRP format 4 3 3 e Definition of input file names via parameters in file PAR Sections 2 2 3 4 2 1 and 4 2 2 e Complete parameter lists with cross reference to the relevant sections in Manual Appendix A and B e Displayed clock s starting time hour and minute can be specified via parameters in PAR file Appendix B e Option for setting all link lengths to that of crow fly distances between upstream and downstream nodes via parameter QCROWFLY in PAR file 4 2 2 5 e For SATURN format network coding various defaults were set in case of implausible coding These include setting link length to the crow fly distance and link free flow speed to 40 km hr if they were coded as zeros 4 2 2 5 e A windows based front end DRACWIN from where the DRACULA programs and the relevant SATURN programs can be accessed DRACULA Manual v2 4 136 APPENDIXES APPENDIX D 2 Changes in DRACULA 2 1 DRACULA 2 1 was released in October 2003 I II MODELLING Improved bus lane changing algorithm to consider the need of overtaking a dwelling bus in order to reach a bus stop in front of Global parameters TINLANE BUS etc to control the cooling period a vehicle keeps after completing one lane changing movement before attempting another one 6 1 7 New junction mo
39. speed of the vehicle to be externally regulated The external activation is achieved by a communication infrastructure in the form of roadside beacons or an autonomous system using on board digital maps combined with a Global Positioning System GPS The main advantage of ISA relative to other forms of urban speed control measures such as 20mph zones or traffic calming measures is its flexibility The system allows for different control speeds at different time of day and different location e g outside schools and during school starting and finishing times and under different traffic roadway and weather conditions The systems were designed initially as a speed management measure for the urban environment However there is no technical restriction of the systems being applied on motorways where they can work in similar manner as controls by variable speed signs such as that implemented on M25 In this latter application the system aims to improve stability and homogeneity of traffic streams in order to reduce interactions conflicts among vehicles which in turn reduce accidents There are a number of variants of ISA systems l Mandatory systems which automatically limit a vehicle s maximum speed to either a prevailing fixed speed limit or to a speed limit varying with road geometry and 2 Voluntary systems which only provide speed limit warnings to the driver such as the system proposed for Borlaagn Sweden TEC 1998 which al
40. the fact that vehicles in the priority flow may take pity on vehicles waiting for a gap and may deliberately slow down in order to create one in which case at least one parameter will be required to indicate their willingness to create gaps This willingness is modelled as a random proportion of priority flow with two proportions FGW BUS and FGW CAR representing the different willingness drivers may show to buses and to other traffic Willingness to create gaps To represent the variable gap acceptance behaviour among drivers Individual DRACULA assigns to each driver a gap acceptance factor You can see from Section 6 1 3 that among the seven vehicle characteristics is a gap drivers gap E acceptance factor and a speed factor These factors represent a kind of risk or acceptance values DRACULA Manual v2 4 85 Section 6 Simulation Settings aggressiveness of the driver a higher speed factor represents a tendency for driving faster a smaller gap acceptance factor the tendency to take smaller gaps hence more risky The safety gap seconds the driver actually uses is the multiple of this gap acceptance factor with an average gap value seconds The reason we specify a gap acceptance factor rather than the actual gap seconds to each driver vehicle is to cater for the fact that the actual gaps required by the same driver may vary from site to site and from time to time 6 1 7 Lane Changing Models Lane c
41. zone number a character C in front of a number denotes a zone See note 1 2 XCORD Its X co ordinate see note 11 3 YCORD Its Y co ordinate NOTES DRACULA Manual v2 4 34 Section 4 Basic Highway Traffic Network and Demand 1 All nodes and zones should be given co ordinates An error 1s registered 1f co ordinates are not given for any node or zone and DRACPREP will terminate un successfully 11 In principle the units used for XCORD and YCORD are arbitrary They are converted to real size in meters by parameter XYUNIT specified in parameter file PAR Table 4 1 However it is strongly recommended that co ordinates are defined in metres and that XYUNIT is set to 1 0 ii Nodes or zones within this section which have not already appeared in the Node or Link records are ignored by the program DRACULA Manual v2 4 35 Section 4 Basic Highway Traffic Network and Demand 4 2 1 5 Turn Priority Marker TPM N B for right hand drive read left for right in this note and vice versa The markers describe which traffic flows oppose a turning movement The naming of TPM follows that of SATURN In addition there are DRACULA specific TPMs The full range of turn priority markers modelled in DRACULA are G Used at a priority junction to indicate the turn which must give way from a minor road As defined in SATURN a G priority marker indicates either a give way or a sharing movements whereby a t
42. 1800F 1 2 34 16 2 1102 105 1103 1 0 103 2 70 200 103 2 4 1103 4 70 200 1800 14 105 2 70 200 1800 1 2 1104 1 0 104 2 70 200 104 2 4 1104 4 70 200 1800 14 105 2 70 200 1800 1 2 105 4 6 85 5 30 101 4 70 200 900S 1 1 1500S 2 3 800S 3 4 102 4 70 200 900G 1 1 1500G 1 2 800G 3 4 103 4 70 200 900S 1 1 1500S 2 3 800S 3 4 104 4 70 200 900G 1 1 15006 2 3 8006 3 4 99999 22222 99999 33333 99999 55555 105 2000 2000 99999 amp SRBS 105 4 6 85 5 30 201 6 2 0 50 40 204 4 2 10 12 4 23 5 2 101 105 DRACULA Manual v2 4 51 Section 4 Basic Highway Traffic Network and Demand 14 8 2 204 201 202 2 2 0 0 40 201 4 2 10 12 4 203 6 2 0 50 40 202 5 2 10 8 4 4 23 5 2 103 105 14 8 2 202 203 204 2 0 0 0 40 203 5 2 2 8 12 4 99999 DRACULA Manual v2 4 52 Section 4 Basic Highway Traffic Network and Demand 4 2 3 3 Lane Choice Modelling at Signalised Roundabouts Rather than representing a roundabout as a series of one way links a distinct advantage of this model is its ability to represent the lane choice of traffic approaching a roundabout Even though a great deal more information is provided to describe such a roundabout and information can be provided for different sections of the roundabout the model treats the roundabout simply as an intersection still Vehicles approaching such a roundabout need only to know which exit from the roundabout to turn as they would know for approaching other types of intersections They would choose an a
43. 4 3 1 1 New vehicle detector record and new responsive signal control logic The detector input includes a new record which specifies the node number of the junction which is linked to the detection It is no longer the downstream node of the link on which the detector is located that is to be controlled It is now possible that a detector can be used to control a junction several links downstream or a junction not even on any downstream routes 8 1 New definition of supply and performance measures for network routes and links Outputs in OUT file 9 1 Link based or approach based gap acceptance values 6 2 1 Roundabout specific circulating speed 6 2 Simplified coding for two way bus route and bus route in reverse order 5 2 1 DRACULA Manual v2 4 134 APPENDIXES I TECHNICAL Generally more checks and report on coding errors New SATURN style data display program SPATULA which takes DRACULA simulation outputs in SPA file and produces a UFD file which can be loaded to the SATURN program P1X for display 9 2 Representation of curved links via SATURN style GIS data input 810 1 Check for non connectivity and banned turns in input routes including bus routes Automatically convert SATURN bus route information in card 66666 into a DRACULA formatted data file in BUS 5 2 1 The name of a bus route can be alpha numeric e g a string of characters and or numbers of up to 5 charact
44. 8 5 1 2 Modelling Public Transport Service The public transport services in the model are described by e service number vehicle type bus or guided bus service frequency veh hr departure time of first service fixed route in terms of nodes through the network a list of bus stops en route Only the departure time of the service via a fixed hourly service frequency is modelled The bus schedule in terms of route timing points will be included in the near future There are two distinct types of buses modelled ordinary buses and guided buses The distinction is made in terms of both vehicle characteristics and with the traffic regulations governing their movement on the streets DRACULA Manual v2 4 67 Section 5 Modelling Public Transport The characteristics for each vehicle are randomly selected from a normal distribution at the start of each trip but are fixed during the trip Section 6 1 3 describes the full range of vehicle types and characteristics modelled in DRACULA and how they can be defined The default average values used for buses and guided buses are listed in the following table compared with those for cars The default coefficient of variation is 0 1 for all parameters Table 5 1 Vehicle characteristics Guided Bus Length m Min safety distance m Normal acceleration m s Normal deceleration m s 7 Max deceleration m s 2 Speed factor O 1 0 1 5 1 6 gt 1 0 0 5 Ga
45. A DRACULA does not accept SATURN sub files using the amp INCLUDE facility Users will have to embed the data into the main NET file 4 2 2 1 Option and parameter specifications SATURN parameter specification records set user defined model parameters for the network They are defined at the top of a SATURN network file under FPARAM The parameters used by DRACPREP relate primarily to those used in the network building and interpretation these parameters are given in Table 4 3 DRAPREP can only read in one parameter per line So if any two of the parameters of Table 4 3 are entered on the same line in a network file one needs to edit the NET file to make them appear on separate lines DRACULA Manual v2 4 43 Section 4 Basic Highway Traffic Network and Demand Table 4 3 SATURN option and parameters used by DRACULA PASSQ A SATURN option TRUE if to have multiple time period assignment AUTOX FALSE If TRUE any uncoded external simulation nodes are automatically coded using the best available data Further testing required Contact the author if you use this function DUTCH FALSE If TRUE node numbers take up to 10 digits EZBUS If TRUE the bus route data on the 66666 data records are in free format By default they are in fixed column format Further testing required FREEXY If TRUE then the X Y coordinates in the 55555 records are input as free format Further testing required LEFTDR TRUE
46. Basic Highway Traffic Network and Demand DRACULA Manual v2 4 66 Section 5 Modelling Public Transport 5 MODELLING PUBLIC TRANSPORT 5 1 The Model 5 1 1 Introduction This section describes the developments made to adapt the DRACULA microscopic traffic simulation model in order to meet the requirements for the modelling of traffic management measures for public transport The work was funded by the UK Engineering and Physical Science Research Council EPSRC as part of a project to look at traffic management issues for kerb guided bus systems Specifically developments have been made in DRACULA in order to model public transport buses and guided buses services reserved bus lanes bus stops and bus laybys passenger volumes and their effect on bus dwell time selective vehicle detection journey time prediction for forecasting the arrival of the bus at the stopline from its upstream detection e a variety of responsive signal control policies such as extension of green early termination of red and payback to maintain offsets in subsequent stages One of the objectives of the EPSRC funded project was to compare guided bus operation with traditional bus priority measures such as reserved bus lanes and bus signal priority The next few sections describe the developments made in DRACULA to represent bus operations and reserved bus lanes Modelling of selective vehicle responsive signal priority measures is described in Section
47. DESCRIPTION l 2 3 NODE NIN JTYPE QYELL NSTAGE or HICD or ZERO OFFSET or NRBL or Node number Number of arms at the node Node type O for external nodes l for priority junctions 2 for roundabouts with or without U turn 3 for traffic signals 4 for a dummy node 5 not in use currently 6 for a new roundabout model see 4 2 3 7 not in use currently 8 for giveway to on coming traffic see 4 2 5 9 for giveway to the right see 4 2 4 a character y or Y entered immediately after JTYPE no space in between JTYPE and QYELL to indicate the junction is in yellow box control See Section 6 2 Number of stages traffic signals only JTY PE 3 or half inscribe circular diameter metres roundabouts only JTYPE 2 or 6 see 4 2 3 or O for other junction type Relative offset traffic signals only or Number of roundabout lanes roundabouts only or DRACULA Manual v2 4 32 Section 4 Basic Highway Traffic Network and Demand ZERO O for other junction type 7 LCY Cycle time for this node seconds for traffic signals or or CSPEED Roundabout circulating speed kph see 4 2 3 and 6 2 or or ZERO 0 for other junction type 8 GAP Junction specific gap acceptance value seconds See 6 2 1 or or BLANK left blank if not to specify tree RECORD TYPE 2 LINK DATA One record for each of the NIN links l ANODE Node at the upstream end of the li
48. DRACULA 2 4 User Manual August 2007 Dr Ronghui Liu Institute for Transport Studies University of Leeds Leeds LS2 9JT Tel 44 0 113 343 5338 Fax 44 0 113 343 5334 E mail trarl its leeds ac uk Web site http www its leeds ac uk software dracula DRACULA Manual v2 4 l Highlights of DRACULA Versions 2 4 This documentation describes the changes made in DRACULA Version 2 4 to be released in autumn 2007 There have been many changes both m the way DRACULA models traffic flow dynamics see the list below on Modelling and in improving the user interface Technical The main changes are listed below MODELLING e Allow a choice of inputting either passenger flow or dwell time at the bus stops Default QPASSFLOW T input passenger flow e Model of time dependent passenger demand profile A same global demand profile is applied to all O D pairs using parameters P NSTEP P TSTEP n and P GONZO n to indicate the number of time steps the period of step n and the matrix factor for step n See Section 5 4 of the Manual e Correction in loading multiple time period assignments from multiple input trp files e Under pre vehicle generation option QNEWDEMAND5 T buses are generated separately from the other traffic because they are on fixed departure time TECHNICAL e Include buses entered both from bus file and SATURN 66666 card in the pre specified vehicle pool when QNEWDEMANDS T is set Use
49. DRACULA Manual v2 4 74 Tip Section 5 Modelling Public Transport 5 2 3 Reserved bus lane data Data in this section must be preceded by a single line containing the string amp BUS LANE and terminated by a single line containing 99999 DATA NAME DESCRIPTION ttk RECORD 1 LINK DESCRIPTION l ANODE Upstream node of the link 2 BNODE Downstream node of the link 3 NRESV Number of reserved lanes on link RAKE RECORD 2 RESERVED LANE l SIDE Relative position of the lane on link measured from kerb side kerbside lane 1 2 RTYPE Type s of vehicles the lane is reserved for 3 XSTART Start position of the lane metre measured from link entry 4 XEND End position of the lane metre measured from the stop line of the link 5 TSTART Start time minutes of the reservation relative to the start of the simulation at the beginning of warm up period 6 TEND End of reservation time period minutes N B the XSTART and XEND are measured relative to different point of the link In DRACULA a lane can be reserved for any type or a combination of several types of vehicle Each type of vehicle has a unique binary number defined by DRACULA See Section 6 1 2 for definition of vehicle types To specify a lane reserved for a single type of vehicle simply give the number of that vehicle type see Section 6 1 2 To specify a combination of more than one type of vehicles use the sum of their vehicle types For
50. DRACULA parameter file par file they will be used for all approaches links 3 If node based gaps are given as in 11111 records they will be used as the GAP and MIN GAP values for links approaching the node and finally 4 if link based gap values are specified in network data files as described in this section these values will be used for the link 6 2 2 Stop line Location By default the stop lines are positioned at a distance of 3 meters upstream from the corner of the road This distance can be specified by parameter X STOPLINE in PAR file which applies to all stop lines in the network It is possible to specify stop line locations for individual links as additional network data in file ADD Enter the records in between markers amp STOP and 99999 as follows DATA NAME DESCRIPTION PERES LINK GAP DATA Eit l ANODE Upstream node 2 BNODE Downstream node 3 XSTOP stop line location metre upstream from the corner of the road In the example net ADD shown in Section 6 2 the stopline on link from node 5 to node 20 lies 5 5 meters upstream from the corner 6 2 3 Reaction zone to junction control A reaction to traffic control at the downstream junction can be specified as a global variable see Section 6 2 or be specified for individual links via input in file ADD Enter the records in between markers amp RZONE and 99999 as follows DATA NAME DESCRIPTION eK JUNCTION REACTION ZONE l ANODE Upstream node 2 BN
51. EASURES eua an a a aa aouadd 117 9 7 DEFINITION OF QUEUE LENGTH oracin a 118 9 8 MULTIPLE SIMULATION RUNS sievccnssssassearsincotesassaeeataericassoecteeassens 118 10 SPECIAL OPTIONS AND FEATURES ccceesececeeereccssssssece 119 10 1 CISD ATA GP MILES eoe T air Cd dr RE AR 119 10 1 1 Enclosed POlyQon icccccccccccccccccseeccccecceeeeseeeeeaaeeseceeeaaeeseesseaaes 119 POT ONES nce a ages as cht DS dala ence da ER 120 LOM GS CONS asas tearing SEO EA CAIR ant nine ee 120 OTR UCM hs et sd pices etal et Anes cae et a 120 10 1 5 Co ordinates of Curved Links 121 10 2 BITMAP BACKGROUND quesareaiscan doam sndsdsaai Sagrada adiantada 121 REFERENCES csomoan aaea a ea ais da 123 APPENDIX A CONTROL PARAMETERS FOR DATA PRE PROCESSING DRACPREP scnis 125 APPENDIX B LIST OF SIMULATION CONTROL PARAMETERS DRAC SIVI saaspeasanesaatera asa oh nar gati san OEE NaN GAVHANE COPAN K N EEEREN RE 127 APPENDIX C THE DRACULA MARS MENUS esseeescsssssseecessesssso 132 APPENDIX D 1 CHANGES IN DRACULA 2 0 cccccccceeeeeeees 134 APPENDIX D 2 CHANGES IN DRACULA 2 1 cccscscseeeeeeees 137 APPENDIX D 3 CHANGES IN DRACULA 2 2 cccscsccceeeeeeees 138 APPENDIX D 4 CHANGES IN DRACULA 2 3 cccsssceoeeeeeeees 140 DRACULA Manual v2 4 6 Section 1 Introduction 1 INTRODUCTION 1 1 The DRACULA Concept The dynamic network microsimulation framework DRACULA Dynamic Route Assignment Combining User L
52. EP Data Type Details of Output Files File extension Default Text reports of the status of the process Network supply description List of route and route flows D Optional List of individual trip departure time route and vehicle characteristics See 4 3 2 The text messages in file TXP contain useful information on the status of the DRACPREP process If DRACPREP fails to complete successfully you TXP SUP EM VEH DRACULA Manual v2 4 9 Section 2 System Requirement and Model Architecture will be prompted with a screen message asking you to check in the file for error messages The file also contains warning messages one particular warning is about the inconsistency between coded link length and crow fly distance between node coordinates Section 4 2 2 2 It is advisable to check the file for any warning and error messages before proceeding on to DRACSIM The data in SUP and DEM are information on the network description and ageregated demand re formatted for DRACSIM They are not meant for users to look at nor to change The outputs in VEA list individual trip s departure time identification of the route it is to take and characteristics of the vehicle A detailed description of its format is given in Section 4 3 2 2 2 2 DRACSIM Function DRACSIM performs the actual simulation of individual vehicles movement in the network following fixed routes Tables 2 4 and 2 5 list the input and output data files f
53. Highway Traffic Network and Demand NOTES 1 M A turn coded M only needs to give way to one stream one lane of traffic from right for driving on the left the stream from the right is called the opposing turn for the M turn Generally an M turn shares the same exit with its opposing turn and they both exit into the nearside lane in exit link In the examples below turn a coded as M gives way to b but not to c By default stream b turns into the same exit link as a In a normally give way situation e g G turn turn a would have to give way to both streams b and c Give way to b not c Exit to lane 1 Eo be CHJ E a Give way to b not c No conflict with d Exit to first lane The merge facility can also be used to model Y junction where two streams of traffic with equal priority or 2 lane link goes into 1 lane There the rule is such that whichever vehicle arrives first to the stop line gets priority DRACULA Manual v2 4 37 Section 4 Basic Highway Traffic Network and Demand 11 S Used on motorway slip road which leads to an acceleration lane a turn marked S has a clear exit lane In the example below turn a enters the acceleration lane without any conflict with the motorway traffic b and c e g does not need to give way to b or c From the acceleration lane the traffic then tries to merge with motorway traffic by changing lane B A Modelled as two junctions at node A and B The slip roa
54. If FALSE right hand drive assumed SPEEDS If TRUE travel speed in kph rather than travel times in seconds are input in the simulation link records NOMADS Duration of the simulated time period in minutes Over writes the TMAIN specified in PAR file Number of user classes to be assigned separately see Section 5 3 2 Pence per kilometre 0 used to convert distances into generalised costs a Ooo Pence per minute converts times into The pcu value per vehicle in user class v or for all vehicles in the trip matrix if unscripted generalised costs VCPCU v XYUNIT coordinates XYFORM The format used to define node coordinates IFGIS n TRUE Control the display of GIS features n 1 2 7 1s for GIS polygon polyline icon text name curved link and node coordinate respectively Currently only options n 1 2 3 and 4 are used by DRACULA DRACULA Manual v2 4 44 Number of metres corresponding to an integer value of 1 as used to define node zone 0 l 1 0 1 0 1 0 I5 Section 4 Basic Highway Traffic Network and Demand In addition certain DRACULA input output filenames can be specified within the SATURN parameter specification record e g record amp PARAM The DRACULA file names can be specified this way are listed in Table 2 6 marked with an An example of filename specification is shown below amp OPTION amp END Otley Basic Network 1985 Example amp PARAM m SPEEDS T
55. If a negative link length is not corrected the simulation will still run with vehicles jumping over the link without stopping However it is advisable to check the report file for any such warnings In many networks particularly SATURN based XYUNIT is left to its default value 1 0 even though the node co ordinates input are not in metres This does not affect the SATURN results as it only uses the link length as defined in node link records However DRACULA simulates individual vehicles second by second movements over the length of the link and then projects the movements onto the physical length of the link in animation It therefore requires the perfect matching of the physical length of a link with the link length as interpreted from the data coded in the simulation node link records Try to make a link s crow fly distance if it is on a straight line or its physical length if on a curved road match with the link length Check file TXP for those grossly incomparable Experience shows that the true value of XYUNIT in SATURN networks is often 10 i e co ordinates are defined to the nearest 10 metres If one wants to use crow fly distance as the link length set the following parameter in PAR file QCROWFLY T In addition DRACPREP sets various defaults in case of implausible coding These include setting the link length to the crow fly distance if it was coded as zero or left blank Similarly a 40 km hr speed is a
56. ODE Downstream node 3 XRZONE distance metre upstream from the junction DRACULA Manual v2 4 91 Section 6 Simulation Settings In the example net ADD shown in Section 6 2 the reaction zone on link 5 20 starts 85 meters upstream to the junction 6 2 4 Roundabout Specification A roundabout is specified through a number of variables HICD Radius of outer boundary of the roundabout NRBL Number of circulating lanes Ristand Radius of the central island Voir free flow circulating speed Section 4 2 2 6 described how some of these variables e g HICD and NRBL are interpreted from information coded for a SATURN format roundabout and others specified as global variables e g Veire and Ristang IN PAR file Now all the above variables can be specified for individual roundabout via additional network data file ADD Enter the data in between record markers amp CIRC and 99999 as DATA NAME DESCRIPTION l NODE Node number of the roundabout 2 CSPEED Circulating speed km hr 3 RISLAND Radius of central island m 4 HICD Radius of the outer boundary m 5 NRBL Number of circulating lanes NOTE l Data for each roundabout should be specified on a separate line 2 Not all four fields for a roundabout need to be specified If a field is left blank or a value 1 is entered the default global value for the variable will be used If a field is left blank all other files afterwards have to be left blank too Otherwise the pro
57. ODES TYPE 4 ROUNDABOUTS TYPE 2 amp 5 SIGHALISED ROUNDABOUTS TYPE 6 amp 7 The simulation speed window shown below allows users to control the simulation speed normally turning the simulation slower so to be able to see the vehicle movements more clearly The window also displays the number of vehicles in process and the number of vehicles from the demand period to be processed The simulation ends when all vehicles from the demand period have completed their journey ES Simulation speed Desired simulation to real time ratio 94 9 15 Current maximum ratio 9999 9 Total no of vehicles in network EE No of vehicles from peak period O DRACULA MARS 1 0 lO x File Disslay Window Smulation Text report 7 Animation Desired sinulainnta rea tireratn 127 pe Current mex mum rato a Total no of vehicles ir retwork Wo of vehic cs from peal periad Simulolivi speed DRACULA Manual v2 4 22 Section 3 Executing the Programs 3 8 2 The Menus Appendix C provides a full of the menu options in DRACSIM 3 8 2 1 File Menu Load network Enable the user to load the network to be simulated Load BMP map To load bitmap background map Screens Enable the user to choose two colour settings 256 colours or 16 32 bits 3 8 2 2 Display Menu Network Info Related information given in the Text Report Window Demand Info Related information given in the Text Report
58. TULA The dialogue SATPIG Generating Route Flows then pops up DRACULA WINDOWS CONTROL Programs Folder Module Run Command Line Run Demos Tools Help 11 Edy a5 ITS ME SATPIG Generating Route Flows Load Network Load Trip Mati ss iis Ses DRACULA Manual v2 4 14 Section 3 Executing the Programs Step3 Click Load Network button to load network UFS and click Load Trip Matrix button to load trip matrix UFM or type the network full path name e g d Dracula net1 ufs and d Dracula net1 ufm into the text boxes W SATPIG Generating Route Flows Load Network Cldraciezeimnb sub UFS Cae RR Fun Cancel Finally start the SATPIG by clicking Run button DRACULA Manual v2 4 15 Section 3 Executing the Programs 3 3 Running the Pre Processor Start DRACPREP from main menu Module Run DRACPREP The dialogue DRACULA DATA PRE PROCESSOR then pops up M DRACULA DATA PRE PROCESSOR Load Network ITS me Click Load Network button to load network or type the network full path name e g d Dracula net1 net into the text box DRACULA Manual v2 4 16 Section 3 Executing the Programs 2 DRACULA WINDOWS CONTROL Programs Folder Module Run Command Line Run Demos headf 12 headg 9 NETI ia HE Type NET File Date Modified 06 09 2001 20 31 i hel Size 2 25 KB Files of type NetFiles Cet Cancel
59. ULA Manual v2 4 87 Section 6 Simulation Settings 6 1 8 Reaction to Traffic Control at Intersections Drivers start to react to traffic controls at intersections when they are within a certain distance of the intersection called junction reaction zone This distance varies with the speed of the traffic if the traffic is moving at a higher speed it will feel the control from the downstream intersection earlier Similarly the lane markings for turning movements may only be displayed within certain distance to the junction where the drivers find they may have to change lane in order to get into the correct lane for their junction turning movements A global parameter is used to represent this effect TAPPRO JNCT a time headway to the downstream intersection when the vehicle will react to junction control and their desired according to their fixed route junction turning movement In the simulation this time headway is then converted into a distanced based junction reaction zone for each link according to the free flow speed of that link RX TAPPRO_ JNCT xV where RX is the distance on link to the downstream intersection whereby the effect of junction is felt by the traffic V is the free flow speed of link In the simulation it is this localised link specific RX which plays an effect see also how to set RX directly for link in Section 6 2 Up the Version 2 2 i e programmes released before 2005 a semi time
60. Window Node Number Show the nodes numbers on the network Bus stop Number Show the bus stop numbers Background map toggle between display or not the bitmap background map GIS ALL Ditto for all of the network GIS information coded in the GIS input data file except for curved link which will always be displayed if given in GIS file GIS polygon Ditto for GIS polygons GIS polyline Ditto for GIS poly lines GIS icons Ditto for GIS icons GIS text Ditto for GIS texts GIS CurvPt Ditto for the points along a curved link GIS CurvPt2 Ditto for the points along and the A and B node number of a curved link Colours A window will pop up which allows changes to the default colour settings for the animation DRACULA Manual v2 4 23 Section 3 Executing the Programs E Colour setting Background 5 Road 7 Clock arms 4 Clock face Lane marking 15 Bus lane marking 11 Blocked lane marking 6 Colour 1 for cars 2 Colour 2 for cars 3 Colour 3 for cars 4 lar 3 Bus 4 Guided bus 4 Taxi HGW 5 Liv 4 1st junction turn O end junction turn 4 ard junction turn 10 e CI F Click on the colour to change and enter the corresponding integer number as defined in Table 3 1 Table 3 1 Colour coding 0 BLACK RED LIGHTMAGENTA YELLOW 7 10 11 12 13 14 5 l DRACULA Manual v2 4 24 Section 3 Executing the Programs 3 8 2 3 View
61. ace between the last node en route and the symbol N B If all or part of a route follows a straight line more or less it is not necessary to define every node along the line but only the nodes at the bends or at the ends The route flow for example network dtest shown in Section 4 2 1 5 are Example NETWORK dtest trp REED E amp ROUTES E LD 200 di 2 3a So 4 P00 25 oe 3 4 4 2 1 100 6 3 4 99999 4 3 1 1 User class vs vehicle type Multiple user classes are used in assignment models such as SATURN to refer to trips which differ with respect to either as defined in Van Vliet and Hall 2002 a vehicle type a their criteria for route choice b network restrictions c demand characteristics e g elasticities For simulation of supply side effects of travel the more relevant distinctions are vehicle types and network restrictions For example lorries and cars have different physical size and acceleration profiles which have an impact on DRACULA Manual v2 4 59 Section 4 Basic Highway Traffic Network and Demand their journey performance Equally cars and taxis would be different 1f there were taxi only lanes In DRACULA the terms user class and vehicle types are often used inter changeably to refer to the physical mechanical characteristics of vehicles and restrictions on them in network access Each user class must be associated with a vehicle type For the definition o
62. adas elis iieia ne 43 A221 Option and parameter specifications cccccceeeeeeeeeeeees 43 4 2 2 2 Simulation Node Data susana anais a andas 45 4 2 2 3 Simulation Link and Turn Data eee 45 4 2 2 4 Simulation Zone Data seara das aceda srs 46 4 2 2 5 Node Coordinates and Link Length eee 46 4 2 2 6 R undaboul amase 47 4 2 5 Signalised Roundabout Model erre 48 4 2 3 1 C odino nSEnChon na E ARE 49 ADD AWE Ee E hc eee 51 4 2 3 3 Lane Choice Modelling at Signalised Roundabouts 53 DRACULA Manual v2 4 4 4 3 DZ 5 3 5 4 5 5 6 1 6 2 7 8 8 1 8 2 4 2 4 Model for Give way to the Right Junction Control 53 4 2 5 Model for Give way to On Coming Traffic 54 HIGHWAY PRAYED DEMAND sas as asilo scarce a A eces sence 58 dos Average ravel Demand Aine ees 58 4 3 1 1 Userclass vs Vehicle typerien ca saida 59 4 3 1 2 Time dependent demand with variable route choices 60 4 3 1 3 Time dependent demand with fixed route choices 61 Has Mamidala a a 62 4 3 3 Route Assignment from SATURN cccccccsseccceceeeseeseeeeeaaeeees 63 4 3 3 1 PCU VS VENC IES e E 64 A332 Multiple user class assignment ccccceceeseeeeseeeeeeeees 65 4 3 3 3 Multiple time period assignment from SATURN 65 4 3 3 4 Removal of sparsely utilised routes cccccccceeecseeeeeees 65 MODELLING PUBLIC TRANSPORT eeererrecccsccscsccsscees
63. additional signal phase when detection of a bus and d plan selection signals which switch to an alternative fixed signal plan when detecting a bus As these signal controls are selective vehicle ones they work hand in hand with the data coding of the detectors described in Section 8 1 The relevant control parameters Table 8 1 in the PAR file Table 8 1 Parameters describing responsive signal controls QBUS a T if there are PT services in the network QPT PRIORITY T if to carry out extension or early a type of priority signal controls EXTENSION Maximum signal extension permitted seconds RECALL 5 The amount of recall time of a red signal seconds NCOOL CYCLE 3 Number of cooling off periods after a signal extension or recall The pane selection control is automatically activated when the appropriate detectors are set up 8 2 1 Journey Time Prediction Two methods are used to predict the time it takes for a vehicle to reach the stopline from its point of detection The first one simply assumes that the detected vehicle will be able to travel down to the stopline at its speed when it is detected The second method considers possible queueing of the vehicle before it reaching the stopline The simple queue model described in Clark et al 1996 is used to predict the DRACULA Manual v2 4 102 Section 8 Advanced Simulation Applications back of the queue at the time of detection using detected lane occupancy infor
64. al vehicles link delay Display link delay wider When display link delay make the link bandwidth wider DRACULA Manual v2 4 133 APPENDIXES APPENDIX D 1 Changes in DRACULA 2 0 DRACULA 2 0 was released in January 2003 and is the first fully 32 bit version I MODELLING Time dependent demand with fixed assignment modelled as step functions 4 3 1 2 Time dependent demand with variable assignment from SATURN multiple time period modelling SATTPX SATSUMA 4 3 3 3 New junction lane choice model giving greater opportunity for vehicles to choose a downstream lane that takes into consideration its next junction turn a looking ahead factor It also considers vehicle types slower moving vehicles made to choose near side lane and vice versa Improved overtaking model for long motorway links after overtaking vehicles prefer to move back Modelling yellow box junctions Section 6 2 Combining rural two lane overtaking using opposite road space with ordinary overtaking over multiple lanes in the same direction 48 5 Definition of vehicle character distributions is extended to include not only means and variances but also upper and lower limits 6 1 New DRACULA style demand input in TRP files a with a new data field to represent user class using the route and b time dependent route flows to be represented in different TRP files 4 3 1 Distinction between user class and vehicle type
65. atory Inputs Detector det Network net Additional Network add Demand trp Pedestrian Network wkw GIS Bitmap gis bmp Discretionary Input Data Simulation Parameter par DRACPREP Status Report Supply Data sup Demand Data dem veh txp DRACSIM Pollution Detector pol Individual Vehicle s Travel Performance ltt E trj Status Report Network Performance txs out reb spa Report txd Bus Performance Passenger Delay pas psn ptt Fig 2 1 DRACULA MARS architecture DRACULA Manual v2 4 8 Section 2 System Requirement and Model Architecture 2 2 1 DRACPREP Function DRACPREP creates the database including the network topology traffic regulation and traffic demand which 1s to be used by DRACSIM Tables 2 2 and 2 3 list the input and output data files for the program Table 2 2 Input data files for DRACPREP Data Type Details of Input Data File Extension Basic inputs Network description See Section 4 2 NET mandatory Travel demand in terms of routes and route flows Section 4 3 Selective Inputs Control parameters Section 4 1 PAR discretionary Public transport services Section 5 2 BUS Selective vehicle detector information Section 8 2 Additional network data inputs Section 6 2 Pedestrian network and flows Section 8 6 Table 2 3 Outputs from DRACPR
66. based variable XAPPRO JNCT is used to describe the junction reaction zone XAPPRO JNCT is defined as the distance metres per every 30 km hr of a speed For example if the link free flow speed is 30 km hr then its reaction zone starts 50 metres by default upstream of the junction If the link free flow speed is 60 km hr then the zone starts 100 metres upstream of the junction The conversion from the variable XAPPRO JNCT to the new variable TAPPRO JNCT is then XAPPRO JNCT 30 x 1000 3600 TAPPRO JNCT 0 12x XAPPRO JNCT and RX 0 12x XAPPRO _JNCT x Vo As long as TAPPRO JNCT is specified for a network in its par file its value will be used in the simulation regardless whether or not XAPPRO JNCT is also specified The response to junction control or lane markings for junction turning restrictions could be in the form of stopping at red signals or in the case of mandatory lane changing drivers become more desperate to change lane 1f they are within such a distance from their target point DRACULA Manual v2 4 88 Example Section 6 Simulation Settings On approaching a give way stop lne and within the distance XAPPRO JNCT drivers would first aim to stop by the stop line When they get closer to the junction and can view the other approaching traffic they could speed up and move off if gaps are available A parameter in DRACULA XVIEW_JNCT is used to describe the distance to the stop line where drivers co
67. cccseseeeeeececseesecesseeeeeeaaaas 92 623 Road Worksand Incident tossine hci tts hb Sonia nadas aliada 92 020 Fellow BOX IUNCIION aids a a O cana 93 O27 Motorway MCF CC siim sa nas ans ass S 93 Oo Flated Approaches qts iiniids iai nda dl eeu 93 EMISSION AND FUEL CONSUMPTION MODELS scsseee 95 ADVANCED SIMULATION APPLICATIONS cccssssseeeees 99 SELECTIVE VEHICLE DETECTORS ines i ENEE E SEENE aS 99 RESPONSIVE SIGNAL CONTROLS cccccccccseccessscessseeesseeeseeeeseees 102 DRACULA Manual v2 4 5 8 2 1 Journey Time Prediction cccccccccccccccssccccccccessecsseceeeasesseeseeaaes 102 O22 SionalExtenstonand Recall ana ss asaa O 103 ds sPOVOGCK apare ieee at a DOR Ta aan D 104 O24 Cooling OM Perod cour assadas ad ara ea a 104 8 2 5 Plan Selection Signal Control 104 8 3 SPEED ADVICE AND INTELLIGENT SPEED ADAPTATION 105 8 4 MODELLING DYNAMIC ROAD PRICING SYSTEMS 107 8 5 MODELLING OVERTAKING ON TWO LANE RURAL ROADS 108 8 6 PEDESTRIAN SIMULATION AND RESPONSIVE SIGNAL CONTROLS 110 9 MODEL OUTPUTS AND EVALUATION ccccsssccoeeeeeees 111 9 1 SUMMARY OF OULU TUBE S ss asian Ss ewes unin uatuieeae sees 111 9 2 THE SPAT UIA PROGRAM assa sussa EN NA 113 9 3 GENERALISED COST assa si otal areas a e 114 9 4 PERFORMANCE VS SUPPLY MEASURES cccccceesceeeseessseensseeuss 115 9 5 EXPEMEASURES Guara codigos Gana Saia DU a e dia a a a 117 9 6 POLLUTION M
68. coding detectors What documented suggests that the kerb lane should be coded as 1 What programmed had treated 0 as the kerb lane Correction is made in the programme e Include in the DRACULA format network coding specification of yellow box junctions e Reset all network limits based on the licensed number of nodes in DRACPREP e Set NGDIGIT to allow user specified digit in outputs spa DRACULA Manual v2 4 139 APPENDIXES APPENDIX D 4 Changes in DRACULA 2 3 DRACULA 2 3 was released in December 2005 I II MODELLING Improved models of roundabout with better management of conflicts at roundabout approaches Add a global variable TAPPRO JNCT to represent the time before approaching a junction when a vehicle begins to react to junction control This replaces the old XAPPRO JNCT which is now used to represent the absolute distance upstream of the junction where the vehicle starts to react to junction control 6 1 Make the equivalent of TAPPRO JNCT and XAPPRO JNCT link specific to be included in the additional data input file ADD 6 2 Start to generate passenger flows to the network after the warm up period This is to ensure that there is sufficient number of buses serving the network and picking up the passengers when the main simulation period begins and in doing so to eliminate the potential problem of excess waiting by passengers at the end of a long bus route Introduce turn priority marker TPM
69. d Eua traffic a coded with an S tum SE E Rm E E E ee eee has a clear exit to lane 1 on gt b lt link A B In the example below a two lane slip road leads to one acceleration lane The on ramp turning movement can also be coded as S The nearside stream of on ramp traffic a will enter the acceleration lane whilst stream b will merge with motorway traffic c as though it has an M turn priority marker A 2 lane slip road Example Ef b merge with c SER SS ia e a Re E a has a clear to lane 1 E a oo a sen DRACULA Manual v2 4 38 Section 4 Basic Highway Traffic Network and Demand The S turn is not restricted to motorway slip roads The left turning movement in the following example can also be modelled as an S turn Whilst stream a has a clear nearside exit stream a will merge with stream b qa 1 lane exit Si A hes tee A oe ek ee eee E Clear first lane exit Exit to first lane only In general an S turn can be used in any nearside turn which has a clear exit lane However an S turn is currently restricted to 2 lanes for the turn e g LANE 1 and LANE2 lt 2 c f Section 4 2 1 2 A far side turn which has a clear exit should be coded as a C turn see Note 111 below For a staged motorway entry see example below each of the two entry streams a and b are best modelled as a separate S turn entering two separate motorway junctions Hence stream b of turn S joins the motorway at
70. d for driving on the right with the turns appropriately mirrored Example ft fie ba An example T junction 2 operating with GWTTR rule DRACULA Manual v2 4 53 Section 4 Basic Highway Traffic Network and Demand Erik Versteegt of TNO provided description to the rules which are involved in this type of junction They are Give way to all e g crossing or merging traffic from the right GWTTR rule Traffic going straight ahead on a road has priority over turning traffic on the same road the X turns Turning traffic with a short corner goes before the long corner corner rule Hence for the above example the priorities are Turn 1 2 4 gives way to turn 3 2 1 based on the X rule Turn 1 2 4 gives way to turn 3 2 4 corner rule Turn 3 2 1 gives way to turn 4 2 3 GWTTR rule Turn 3 2 1 gives way to turn 4 2 1 GWTTR rule Turn 4 2 3 gives way to turn 1 2 3 GWTTR rule Turn 4 2 3 gives way to turn 1 2 4 GWTTR rule The X turn can already be modelled in DRACULA The corner rule is implicitly modelled in DRACULA with the traffic from the short corner getting onto the junction quicker than those from the longer corner 4 2 5 Model for Give way to On Coming Traffic Often seen on rural country roads are sections of the road which are too narrow to accommodate simultaneous crossing of vehicles from both directions For example the narrow bridge Sutton Bridge pointed by t
71. d links from the downstream node For each permitted turn the lane s in the link that can use this turn are specified and a marker describing its priority over opposing flows is given A lane can be reserved for a particular type s of vehicles only for example a reserved bus lane This section defines coding of a basic highway network which has no reserved lanes the latter is described in Section 5 4 Zones and nodes are named with a number There should be a unique number for each zone and a unique number for each node The numbers do not need to be sequential In addition nodes and zones may have the same numbers DRACULA network coding adopted a convention used in SATURN to distinguish a zone number from a node number by including the character C with the zone numbers A NET file may use one of two different coding conventions a that used by SATURN networks Section 4 2 2 and 2 a DRACULA specific set of Ti records 4 2 1 Selection of a is made according to parameter P QSATNET T to be specified in PAR file Note that the SATURN formatted files contain a lot of data that is not required by DRACULA and equally do not contain all the information which DRACULA requires Nevertheless they offer a convenient starting point for users with existing SATURN networks 4 2 1 DRACULA Network Coding To describe a basic highway network there are five sections of records records 1 and 5 are discretionary whilst records 2 4 are man
72. d to approximate the traffic composition Based on the proportions of each type of vehicles within the overall demand as specified by parameters in PAR file Table 6 3 the model randomly assigns a type to each individual vehicles according to a probability distribution A demand factor via parameter GONZO is used in DRACULA to change the overall level of demand This is useful when simulating the effect of different traffic demand levels with fixed route choice on networks The name of this parameter is borrowed from SATURN where the GONZO value affects not only the level of demand but also in congested networks the route assignment results Table 6 3 Parameters used to specify traffic composition and demand factor Defaults i ee PCU value for the overall demand PBUS Proportion of total demand is buses For 10 enter 10 PTAXI Proportion of total demand is made up by taxis PHGV Na Proportion of total demand is HGVs Proportion of total demand is LGVs DRACULA Manual v2 4 83 Section 6 Simulation Settings For example 1f GONZO 1 5 and VCPCU 1 2 the program DRACPREP will first multiply the whole demand by these two factor and then assigns the proportions PBUS PTAXI PHGV PLGV of the resulted demand to be of type BUS TAXI HGV and LGV The rest is of CAR type The proportions should be specified such that if PBUS 10 it means 10 of the total demand is of type BUS Clearly the total should b
73. datory Each section is preceded by a line containing a string beginning with character amp and followed by a section name and terminated by a line containing a single record 99999 The five sections are l Section 1 File name description Preceded by amp FNAME DRACULA Manual v2 4 30 Section 4 Basic Highway Traffic Network and Demand 2 Section 2 Node and link description Preceded by amp LINKS Within this Section there are three types of records 1 Record type 1 Node description mandatory One record for each node 2 Record type 2 Link description mandatory One record for each link in strict clockwise order but starting with any arbitrary link 3 Record type 3 Traffic signal stage description Only required for traffic signalised nodes One record per stage 1s input 3 Section 3 Description of zones Preceded by amp ZONES One record per zone 4 Section 4 Node zone coordinates Preceded by amp COORD One record per zone node coordinates 5 Section 5 Partially signalised roundabout Preceded by amp SRBS See Section 4 2 3 for details Coding instructions and a worked example for a basic network are given in the following sub sections The NAME provides an convenient shorthand for data fields for later discussion while DATA column records the number of fields for the record Data fields are free format there is no column
74. dels for merge motorway slip road and turning movements with clear exit 4 2 1 5 New type of junction control always give way to traffic from left or right for driving on the left or right This is a type of junction control frequently seen in Dutch networks and some networks in Belgium 4 2 4 TECHNICAL Options to load bitmap background map in BMP format similar to that used in SATURN 10 2 From Version 2 1 SATURN format GIS files can be loaded directly into DRACULA without any modification In previous versions the corners of a polygon poly line or a curved link have to be specified manually in gis file See 10 1 Bitmap and GIS file names can be specified via parameters in either the network file net or the parameter file par Alternatively a GIS file can also be picked up by DRACSIM automatically if it shares the same name as the network name 2 2 3 The bitmap file can be loaded via command line input 3 6 or via DRACSIM file menu Appendix C Include SATURN program SATPIG within DRACWIN front end for the convenience of DRACULA users SATPIG extracts route assignment results from SATURN in a format directly importable to DRACULA 33 Allow removal of routes which have very small flows 4 3 3 4 DRACULA Manual v2 4 137 APPENDIXES APPENDIX D 3 Changes in DRACULA 2 2 DRACULA 2 2 was released in December 2004 I II MODELLING Allow mixture of two lane rural roads where
75. der development TWARM 10 Warmup period minutes Section 6 1 1 DRACULA Manual v2 4 131 APPENDIXES APPENDIX C The DRACULA MARS menus Load network Load in simulation network Load BMP map Load in bitmap background map in bmp format Exit Close graphics window first by clicking the OK button before exit Screen Change screen colours Display Network info Display network information Display demand description Zone number Toggles to display or not the zone numbers Node number Toggles between display or not display node numbers As above for bus stop numbers Background map As above for background map GIS data As above for GIS data Full network Fit the full network on screen Full background Fit the bitmap according to either the original coordinates specified in XYB file or to align the image to bottom left corner of screen Windows Select to change either network bitmap or both Use left hand muse button to select the area for viewing Zoom Zoom in omens Move the current viewing box to the left DRACULA Manual v2 4 132 APPENDIXES Toggles between with and without graphical animation Display vehicle by O Display vehicle by vehicle type Junction turning by vehicle s downstream junction turning movement by routes for limited number of routes only Type amp randomly By vehicle type and randomly select colours for cars Animation by ns Display individu
76. ding files EXTENSION Only the names of the files need to be entered DRACPREP and DRACSIM will automatically pick up the relevant extensions For example to specify otley BUS set BUSFIL otley The names can also be put inside inverted commas as in SATURN coding e g for the above example BUSFIL otley Table 2 6 Parameters used to specify the file names period t Section 4 3 1 2 This facility is useful to allow the same input file be used in a number of different scenario testing For example the same demanded trips OTLEY TRP can be used as basis for two different network descriptions OTLEY1 NET and OTLEY2 NET DRACULA Manual v2 4 12 Section 2 System Requirement and Model Architecture N B Before DRACPREP can load in the correct network coding format see the two different ways of coding networks in Sections 4 2 1 and 4 2 2 the relevant parameter needs to be specified in PAR file Hence the PAR file has to bear the same name as that of the NET file As an experiment those filenames marked with marked can also be specified in the network description file NET PARAMETERS PERIOD MIN 60 NSEED 900 LEFTDR T QBUS T FILTRP otley trp FILBUS otley bus FILADD otley add FILDET otley det END DRACULA Manual v2 4 13 Section 3 Executing the Programs 3 EXECUTING THE PROGRAMS Programs within the DRACULA MARS suite can be accessed using either the DRACWIN Section 3 1 or the SATWIN Secti
77. dynamic traffic demand management measures such as congestion pricing Through its animation DRACULA MARS gives a graphical representation of the network performance 12 Overview of DRACULA Traffic Simulation Model The traffic model in DRACULA is a micro simulation of the movement of pre specified vehicles through the network Drivers follow their pre determined routes and en route they encounter signals queues and interact with other vehicles on the road A large number of such microscopic vehicle models have been developed in the past at varying levels of complexity and network size e g in some the network is effectively a single intersection An essential property of all such models is that the vehicles move in real time and their space time trajectories are determined by e g car following and lane changing models and network controls such as signals The simulation is based on fixed time increments the speeds and positions of individual vehicles are updated at an increment of one second Spatially the simulation is continuous in that a vehicle can be positioned at any point along a link The model includes an animated graphical display of vehicle movements in the network The simulation starts by loading the simulation parameters network data including global and local variations and trip information demand and routes determined by the demand model It then runs through an iterative procedure at the pre defined time incre
78. e less than 100 e g PBUS PTAXI PHGV PLGV lt 100 Indeed they should be much less than 100 in most of cases as the traffic is composed mostly by CAR 6 1 5 Car Following Models Car following models represent the longitudinal interaction between vehicles The speed of the following vehicle is modified in the light of the relative speed and position of the preceding vehicle The parameters required to determine such longitudinal progress of vehicles are the desired speed the desired minimum headway the reaction time the rate of acceleration and the rate of deceleration The implementation of such parameters in DRACULA is via text input file VEH TAB details are described in Section 6 1 3 6 1 6 Gap Acceptance Models Gap acceptance models are used to describe how drivers make the decision to merge or cross at intersections They are so called because they represent the critical gap seconds in the opposing stream s of traffic that the driver feel safe to accept if the gap is greater or equal to the critical gap then merge or cross otherwise wait The key parameter for gap acceptance models is the normal acceptable gap expressed in seconds for the manoeuvre being contemplated Some gap acceptance models use a fixed value for each manoeuvre DRACULA allows the minimum gap to be reduced if the traffic is very heavy and moves slowly a representation of frustration on the part of the waiting traffic DRACULA uses the time a driver ha
79. earning and microsimulation has been developed at the University of Leeds since 1993 Liu et al 1995 It adapted a new approach to modelling road traffic networks whereby the emphasis is on the microsimulation of individual trip makers choices and individual vehicles movements The model attempts to represent directly the behaviour of individual drivers and vehicles in real time as these evolve from day to day This is coupled with a detailed within day traffic simulation model of the second by second movements of individual vehicles according to car following lane changing rules and traffic controls In combination they model the evolution of the traffic system over a representative number of days so that both within day and between day variabilities are included and interaction between the demand and supply modelled The full DRACULA framework combines a number of sub models The demand model represents the day to day variability in total demand within a fixed departure time period It simulates for each potential traveller based on individual drivers knowledge of the network their past experience and perceived network condition of the day whether to travel if so the route to be taken and the preferred departure time This information is then passed to the traffic simulation model which represents the within day variability of network conditions and simulates individual vehicles movements through the network following the rou
80. ececececesssssssscsseeeece 13 DRAC WIN FRONT END sarro na n sa astan Ganda sda a nas anca sh dn sais n assa 13 RUNNING THE SATPIG PROGRAM ccceeceeeeeeeeeeeeeeeeeeeeeeeeeeeeeees 14 RUNNING THE PRE PROCESSOR siruina 16 RUNNING THE SIMULATION aro ai ai ici 19 RUNNING THE SPATULA DATA PROCESSING cccceeeeeeeeeeeeees 20 ACCESSING THE PROGRAMS FROM SATWIN FRONT END 20 RUNNING THE PROGRAMS IN BATCH MODE 20 THE DRACSIN UW INDOW rretra i nET A 21 3 8 1 The Winds rss fd dd iactuncsewaet 21 3 8 2 TEM fe dh A RR RR RE RR RR RS 23 3 8 2 1 EO CNC rea Ra A aaa 23 3 06 22 Dasplas MN sea ia 23 3 8 2 3 CNIC ol VIC INU sai in ia sa ni 25 3 8 2 4 Simulation INICIO esco saias o ra ae rias a GR asd 25 BASIC HIGHWAY TRAFFIC NETWORK AND DEMAND 27 CONTROL PARAMETERS FILE eisena ee asso si sass ganda 27 BASIC HIGHWAY NETWORK ie dio ELOS ERR GOO et Merete a 30 4 2 1 DRACULA Network Coding ccccccccccccccsssseeesscececcseasesesseeeeeenes 30 4 2 1 1 File Name SDECINICALION asas ai canis es ia neues es deaousen 31 4 2 1 2 Node and Link Data Formats cccccccccccccseeeeeeeeees 32 AD eld POMC TIAA BOTAS ean TN N 34 4 2 1 4 Node and Zone Coordinates 20 cccccessssssessseeeeeeeeeeeees 34 4 2 1 5 T m Priority Marker TPM asas a ei ieee 36 4 2 1 6 Link Length and Junction Layout 40 Ao Aea Co res aa RR ear ae 42 42 2 SATURN Network Coding meias asnsaiondecR
81. ed The section should be preceded by amp SRBS and ending with a string 99999 This section of data contains three mandatory record types Node description as in the basic network coding 2 Description of approach arm to the roundabout 3 Sections on roundabout DRACULA Manual v2 4 48 Section 4 Basic Highway Traffic Network and Demand 4 2 3 1 Coding instruction For DRACULA networks the DATA column specifies the number of fields for the record As with the DRACULA network coding convention data fields are entered in free format and only a white space is required to separate two fields For SATURN networks one need to look at COL which indicate the precise column s where each data field needs to be entered DATA COL NAME DESCRIPTION 1 1 5 NODE Node number 2 6 10 NIN Number of links at the node 3 11 15 JTYPE Node type 6 for the new roundabout model 4 16 20 HICD half inscribe circular diameter metres JTY PE 6 only 5 20 25 NRBL Number of roundabout lanes Note 1 6 26 30 CSPEED Roundabout circulating speed kph meee RECORD TYPE 2 1 ROUNDABOUT APPROACH DATA 1 6 10 RBNODE Node number at the approach Note 111 2 11 15 RJTYPE Node type 1 for priority approach 3 for signalised approach 3 16 20 RNSTAGE Number of stages if RJTY PE 3 or or 0 if RJTYPE 1 4 21 25 ROFFSET signal offset if RJTYPE 3 or or 0 if RJTYPE 1 5 26 30 RLCY cycle time if RJTYPE 3 or or 0 i
82. eded by amp ALT SIGNAL and terminated by 99999 Within this section are two records 1 Record type 1 Signalised intersection description One record per node 2 Record type 2 Signal plan description One record per signal plan The coding format 1s DRACULA Manual v2 4 99 Section 8 Advanced Simulation Applications DESCRIPTION eR SECTION 1 DETECTOR DATA DATA NAME l DETID 2 QDEM 3 VTYPE 4 QSTAR 5 NODE 6 ANODE 7 BNODE 8 SIDE 9 POSIT 10 RECALL or RDETID 11 TCOOL Detector number A character to be entered immediately after DETID indicating the type of responsive signal control to trigger if QDEM d or D to trigger demand dependent signal controls described in Section 8 2 2 if QDEM or T coupled with a detector with QDEM r or R to trigger a plan selection signal as described in Section 8 2 4 Default QDEM BLANK the detector is not used to trigger any responsive signal control but maybe used to collect data at the detectors Types of vehicles to detect A character to be entered immediately after VTYPE to indicate how the detector location is measured If QSTAR POSIT is measured relative to the stop line of the link By default QSTAR is left blank where the detector location is measure relative to link entry Node number of the signalised junction to be controlled Upstream node number of the link where the detector is located Downst
83. eed to define any PCU value for buses to convert SATURN bus data to DRACULA bus data 4 3 3 2 Multiple user class assignment To import results from a SATURN multiple user class assignment to DRACULA set the corresponding PCU values Section 4 3 3 1 and vehicle types for each user class 4 3 1 1 in PAR file 4 3 3 3 Multiple time period assignment from SATURN The quasi dynamic assignment modelling of SATURN models time dependent matrices and the effect of over capacity queues on time period assignment See Chapter 17 of the SATURN Manual on how to set temporal demand in DAT file and how to run SATTPX and SATSUMA to produce multiple time period assignments The result is a series of SATURN UFM and UFS files one pair for each time period Running SATPIG on each pair of the UFM and UFS files the SATURN route choice for each time period is extracted to DRACULA format TRP files Then following the instructions given in Section 4 3 1 2 and Table 4 5 to feed these data into DRACPREP 4 3 3 4 Removal of sparsely utilised routes Some routes generated by SATURN have very small flows To remove such routes from DRACULA simulation the following global parameter can be used to set a limit on the absolute minimum flow a route a selected by DRACULA Table 4 8 Parameter to limit small flow routes FIJPMIN Minimum flow veh hr below which a route is not selected for DRACULA simulation DRACULA Manual v2 4 65 Section 4
84. emand for example an average demand level DRACULA Manual v2 4 19 Section 6 Simulation Settings A warm up period T is simulated with demand linearly increasing from a fraction F of the peak level to the peak as a way of ensuring that the traffic from the demand period does not start with an empty network At the end of the main period the simulated demand linearly decreases from the peak level to a level F equivalent to that of an off peak over a cooling off period Tc and thereafter stays at that level until the end of the simulation tena Various time periods and associated demand levels shown in Fig 6 1 can be specified via input parameters in PAR file Symbols Parameters Defaults ee TWARM 15 min By FWARM 0 5 see Note T TCOOL 15 min Fe FCOOL 0 5 see Note Ta TMAIN 60 min NSTEP F FSTEP 1 F FSTEP 2 etc T TSTEP 1 T TSTEP 2 etc Note For historic reasons the warm up and cooling of flows are modelled as a proportion to that of the peak demand Hence a Fw 0 5 means that the demand during the warm up period will increasing linearly from half of the peak demand of that of the first time period to the peak demand Similarly over the cooling off period the demand will be reduced linearly from the peak demand to by default half the peak level F 0 5 By default a simulation run in DRACULA ends when all vehicles departing during the demand period have finished their journey The
85. end of simulation is also indicated with the No of vehicles from peak period on the Simulation Speed screen becoming zero The actual simulation time will then depends on the level of congestion in the network and the lengths of the journeys For example if the longest journey in a simulation takes 10min to complete in free flow situation one would expect the simulation will terminate in 15 20 minute s with moderate congestion in the network It is possible however grid lock may occur in the simulation In this case the simulation could run forever To eliminate such situation the simulation will only run for a total of three times of the demand period counting after the warm up period after that period the simulation will terminate and an error message will be reported Such situation may also occur if the average journey time is longer than three times of the demand period specified DRACULA Manual v2 4 80 T ip Section 6 Simulation Settings It is possible to terminate the simulation at the end of the main demand period i e after TWARM TMAIN This can be done by setting the following parameter value in the par file QMAINSIM T The choice between running a simulation over a pre defined time period and tracking all vehicles from the demand period to their destinations and therefore an unspecified time period depends on the simulation outputs one wants please see Section 9 4 for a distinction between the
86. ent metre Section 8 5 The display colour of dummy vehicles used to mark a blocked lane Default black colour The 16 bit colour palette is used the numerical coding of the colour is defined as follows 0 BLACK 1 BLUE 2 GREEN 3 CYAN 4 RED 5 MAGENTA 6 BORWN 7 GREY 8 DARKGREY 9 LIGHTBLUE 10 LIGHTGREEN 11 LIGHTCRYAN 12 LIGHTRED 13 LIGHTMAGENTA 14 YELLOW IS5 WHITE Ditto for cars COL DUMMY COL CAR 3 COL GBUS Ditto for guided buses COL TAXI Ditto for taxi COL LGV Ditto for light goods vehicles DRACULA Manual v2 4 127 APPENDIXES COL HGV EXTENSION Fl OVERTAKE F2 OVERTAKE Ditto for heavy goods vehicles Maximum signal extension permitted sec Section 8 4 Speed improvement factor Section 6 1 7 Speed constraint factor Section 6 1 7 Start level of traffic flow for the cooling off period as a proportion to the peak demand Section 6 1 1 FCOOL FGW BUS FGW CAR FIJPMIN fraction of traffic willing to give way to buses Section 6 1 6 Courtesy factor fraction of traffic who are willing to give way to minor flow cars or cars wanting to change lane Section 6 1 6 Absolute minimum flow veh hr en route Section 4 3 3 4 Average acceptable gap seconds Section 6 1 6 GAP MIN Minimum acceptable gap seconds Section 6 1 6 GAP TSTART GAP TEND GONZO IFGIS n LEFTDR F if driving on the right e g European driving Waiting time sec before reducing
87. entification location on the link kerb or median lane type s of vehicles reserved for the lane set backs at the begining and end of the link start and end time of the reservation measured from the start of the simulation It is assumed that bus drivers all have perfect knowledge of the network Thus they know if there is a reserved bus lane in the next link well in advance and will try to move into a lane in the current link which leads it naturally geometrically into the reserved lane in the next link If it is not possible a bus would go into the next link first then keep looking for chances to move into the reserved lane as long as the reserved lane permits its next junction turning or else it is not too close to the end of the reservation In contrast a private vehicle driver only obtains information about their next link when they approach the junction If the start of the reserved lane in the next link is further than 50 metres from the upstream end of the link the driver would ignore the reservation and choose to get into a lane in the next link according to its position in the current link and the traffic situation in the next link If the set back of the reservation at the beginning of the next link is less than 50 metres long the private vehicle driver would choose a lane from the remaining lanes in the next link merging with other traffic from the same link if necessary Moving in a link containing a reserved b
88. ere net ufs Input post assignment network file trips ufm Input trip matrix net TRP Output route file The outputs file net TRP are in the format described in Section 4 3 1 and hence can be loaded in by DRACULA straight away In versions earlier than SATURN 10 3 the route flows were stored in file net LPG which combines SATURN line printer and route file The format of LPG files are very different from that of TRP To enable DRACPREP to load route flows from LPG format set the following parameter in PAR file QSATPIG T DRACULA Manual v2 4 63 Section 4 Basic Highway Traffic Network and Demand 4 3 3 1 PCU ys vehicles Flows in SATURN are measured in pcu hr they need to be converted to vehicles hr A conversion factor PCU can be specified in PAR file for DRACPREP to carry out the conversion for a single user class assignment or single type of vehicles simulated For multiple user class assignment each user class should have a specific value of PCU factor input in PAR file as indicated by a subscript Thus PCU 2 1 5 would assign a pcu factor to user class 2 Table 4 7 lists the parameters used to specify user class dependent PCU values Table 4 7 Specification of PCU values for single and multiple user class nal PCU value per vehicle for all vehicles in the trip matrix VCPCU v a PCU value per vehicle for user class v When loading demand from SATPIG outputs DRACPREP will convert SATURN flow in pcu h
89. eriod assignment only Results for multiple time period assignment can be load into DRACULA via separate TRP files each file contains the route assignment for each time period Table 4 5 lists the essential parameters used to specify a temporal demand distribution and file names of route assignment for each time period DRACULA Manual v2 4 60 Section 4 Basic Highway Traffic Network and Demand Table 4 5 Specification for a temporal demand distribution with time dependent route choice NSTEP l Number of time steps Maximum 10 steps TSTEP 1 da Duration minutes for the first time period TSTEP t a Duration minutes for time period t t 2 3 10 T Set it to be T to load time dependent route choice from files TRPNAM FILTRP t File name with extension TRP of route choice for time period t t 1 2 10 4 3 1 3 Time dependent demand with fixed route choices It is possible to model simply a temporal demand profile without route re assignment 1 e with a fixed route assignment for all time periods In this case only one TRP file is required which contains the fixed route assignment and the file name can be specified via parameter FILTRP or FILTRP 1 in the PAR file Table 4 6 lists the parameters to be specified in PAR in order to specify the temporal demand distribution with fixed route choice Table 4 6 Specification for temporal demand with fixed route choice Defaults Description NSTEP l Number
90. ers 5 2 1 If all or part of a bus route follows a straight line more or less it is not necessary to define every node along the line only the nodes at the bends This applies to the routes that are manually created in a TRP file New bus service record in BUS file does not require input of origin destination zones This is in recognition the fact that bus services may not always begin end in a conventional zone The old format where origin and destination zones for a bus service are coded can still be used by setting parameter QOLDBUS T in PAR file 5 2 1 A population vehicle list created by DRACPREP in VEH file which contains all vehicles to be simulated and their departure times and characteristics This can be used as a fixed population from one simulation to another 4 3 2 Automatically convert SATURN junction based gap acceptance parameters GAP GAPR GAPM to DRACULA style link based gap acceptance values 46 2 Read m SATURN network coordinates using the DUTCH format Options in Display menus of DRACULA MARS Window to change colour schemes used in simulation including those to colour a bus lane and a blocked lane 3 7 2 Appendix B Check coding on lane usage whether all lanes being used for turns All additional network descriptions are to be coded in file ADD Additional coding currently includes junction dependent gap acceptance parameters 6 2 1 roundabout circulating speed 6 2 rural
91. erts the driver if they infringe a speed limit and registers a violation if the warning is ignored Driver compliance can be achieved either through the normal use of the vehicle control the advisory systems or by driver enacting the change say by pressing a button to allow the vehicle s speed be regulated to the speed limit the driver selection systems For modelling purposes a more useful distinction is the actual number or percentage of vehicles which are under or are complied to ISA control and the speed limits on each section of the road Later in the section when reference is made to ISA penetration rate 1 e the percentage of vehicle fleet equipped with the control devices of the mandatory systems it could equally referred to the level of compliance in voluntary systems The speed limits are inputs to the simulation modelling there can be more than one speed limit sections on one link which allows modelling a smaller speed limits at road bends for example Once specified at the start of a simulation the speed limits do not change over the simulation period To enable speed control to be modelled enter the relevant link speed limits in the additional network data file ADD between amp ISA and 99999 delimiters The speed control may not apply to all parts of the network so DRACULA Manual v2 4 105 Section 8 Advanced Simulation Applications one needs only to enter the links that are under the speed contr
92. example a lane which has a reserved type 2 is for bus only 6 for bus and guided bus 10 for bus guided bus and taxi A lane reserved for type 63 is for all types of vehicle e g not reserved and O means it is a blocked lane One can therefore model a local incident that causes partial lane closure for a specified period of time DRACULA Manual v2 4 15 Section 5 Modelling Public Transport 5 3 An Example The following example public transport data inputs show that there are two public transport routes in the network both served by buses the first service number 30 begins its operation 10 minutes after the start of the simulation period and runs every 10 minutes along a route that passes 17 nodes and stops at 4 bus stops the service number 33 does not stop anywhere en route it is probably an express bus there are two reserved lanes in the network they are along links 1511 721 and 721 1511 Both reserved for buses with setback of 150 metres and operate all day amp BUS SERVICE 30 2 10 10 7 FOU PO DD TOS Hom GAI TRI RD D gt 5 13 21 9903 9908 132 182 9910 Example 4 101 102 104 107 rae de Ra a 12 773 777 782 12 15 13 21 9903 9908 132 182 9910 0 99999 amp BUS STOP 101 722 723 100 012025 102 15 13 500 035 104 219903130 00 020 107 132 182 90 012025 99999 amp BUS LANE 1511 721 1 1 2 0150 036000 721 1511 1 1 2 0150 036000 99999 5 4 Guideway for Guided Buses A guideway is generally thoug
93. f RJTYPE 1 6 35 40 RSHIFT entry and exit separation meters See Note 11 fee RECORD TYPE 2 2 ROUNDABOUT LINK DATA 1 10 15 RANODE Upstream node of the roundabout section 2 15 20 RTIM Free flow speed km hr on this section 3 20 25 RDIST Length of the section m Not in use 4 26 30 RLANES Number of lanes of the section Note 1 5 3 1 35 RLMARKI Lane marking for the first lane 6 36 40 RLMARK2 Lane marking for the second lane 7 40 45 etc etc for all RLANES DRACULA Manual v2 4 49 Section 4 Basic Highway Traffic Network and Demand We RECORD TYPE 2 3 SIGNAL DATA AS IN SATURN 1 10 15 2 15 20 3 20 25 4 26 30 5 31 35 6 36 40 7 40 45 8 45 50 Note 1 1 iii iv DRACULA Manual v2 4 STAGL Duration of stage seconds INTG Duration of following inter green seconds NGM The number of node entries which follow as GNA 1 GNC 1 GNA 2 GNC 2 with two entries per green movements GNA 1 The A node for the first green movement GNC 1 The C_node for this turn GNA 2 The second A node GNC 2 The second C_node etc etc Different sections of a roundabout could operate with different number of lanes NRBL should be the maximum number of lanes on the roundabout while RLANES the number of lanes along specific section NRBL is used in this signalised roundabout model purely for display purpose the roundabout will be shown with equal number of lanes In the simulati
94. f vehicle types used in DRACULA see Section 6 1 2 More than one user class or vehicle type may use the same route It can be noted that user classes or vehicles types are embedded within the coding of a route If a route is used by more than one user classes each user class should be entered with the respective MUC specified with MUC refers to the first second and the n th of user class using this route By default all user classes are assumed to be of vehicle type CAR To specify different vehicle types use the following parameters in the PAR file Table 4 4 Specification of user classes NMUC l Number of user classes in the demand input Maximum 10 user classes MUC v 1 Vehicle type for user class v See Section 6 2 for numerical representation of vehicle type Type 1 vehicles are CARS 4 3 1 2 Time dependent demand with variable route choices Temporal distribution of demand can be represented in DRACULA by step functions indicating the level of demand for each time period step and the route assignment for the time period N B Time periods discussed here are sections of time within the main demand period To conduct a simulation of the main period a warm up time period before and a cooling off period after the main period are also simulated The warm up and cooling off periods are described in Section 6 1 1 It should be noted that the coding of routes in DRACULA permits input of Ti a single time p
95. fic Network and Demand NOTES 1 If any exit node GNC above is zero it is assumed that ALL turning movements from that entry node GNA except of course any prohibited movements indicated by zero turning lanes LANE and LANE2 are allowed 11 Code all movements in one stage on the same line 11 In versions priori to January 2004 version 2 1 at most two links are allowed in any one stage There could be as many turns from the two links as required in each stage This restriction has now been removed 4 2 1 3 Zone Data Formats Data in this section must be preceded by a single line containing string amp ZONES and terminated by a single line containing 99999 DATA NAME DESCRIPTION 1 ZONE Zone number 2 NODEA Upstream node of the link connecting the zone 3 NODEB Downstream node of the link connecting the zone N B Strictly speaking node NODEA should be of type external where traffic enters to and leaves from the network In SATURN formatted networks Section 4 2 2 it is possible that a zone is connected to an internal node In that case DRACULA treats the node as though it is external such that no traffic control applies to traffic entering leaving the network at that node 4 2 1 4 Node and Zone Coordinates Data in this section must be preceded by a single line containing string amp COORDS and terminated by a single line containing 99999 DATA NAME DESCRIPTION l NODEN Node or
96. g seasonal bus tickets and b the time for the door of bus to be opened and closed A value of a 4 second passenger and b 5 second are suggested Clark et al 1996 Different values can be specified by the users via parameter setting see coding in Section 5 2 In the model a bus has unlimited passenger capacity Passenger routes are not modelled nor is delay due to passengers egress Once a bus stops at a bus stop it will pick up all the waiting passengers regardless of the number of passengers and regardless of where they want to go Consequently when a bus stops at a bus stop no other buses need to stop at that bus stop A bus only stops if there are passengers waiting While a bus is picking up passengers at a bus stop any more passengers arriving at the bus stop will be added to the passenger queue and the bus dwell time will be prolonged accordingly However once a bus is in the mode of closing its door or has closed its door 1t would not re open its door for the new arrivals In the case of a bus layby even if the bus is still waiting to leave the layby passengers newly arrived at the layby still have to wait for the next bus The next bus sees the waiting passenger and will aim to stop at the layby which helps to create gaps for the waiting bus to leave the layby DRACULA Manual v2 4 69 Section 5 Modelling Public Transport 5 1 4 Modelling Reserved Lanes Reserved lanes are specified in the model by e the link id
97. gestion Metering system is that charges would only be levied when delays occur This is possible by the setting of a congestion threshold in the form of a critical time value for ntersecti a certain unit of distance above which a charge would be levied The suggested threshold in Cambridge was three minutes to travel half a kilometre equivalent to an average speed of 10kph Oldridge 1990 When the threshold 1s exceeded the charge levied may be a pre specified fixed unit a figure of 20 Pence Sterling was mentioned in Cambridge or a variable unit that depends on the degree of threshold violation e g a charge in Pence Sterling per minute for all time taken above the threshold Although calculation of the threshold is continuous i e carried out at much more frequent intervals than the half ntersect distance unit charges cannot overlap so that a fee will only be levied once for using any particular half kilometre stretch of congested road Previous modelling work to investigate the route and demand choice aspects of a range of charging technologies in a static modelling context has been forced to rely on a much coarser specification of the Congestion Metering system May and Milne 2000 A faithful representation can only be achieved using a fully microscopic simulation model where costs may be applied separately to individual vehicles rather than to aggregate flows In particular a feature of the threshold system is that the points in space
98. gram will read the field after as that meant to take the default The example shown in Section 6 2 defines for roundabout of node number 1115 a circulating speed of 20kph and leaves all other variables their default values The example also specify for roundabout 1116 a circulating speed of 30kph radius of the central island that of the default value outer radius of 20 2 meters and with 4 circulating lanes 6 2 5 Road Works and Incidents DRACULA can model planned road works and regular incidents such as illegally parked vehicles where location and duration can be specified Such incidents are modelled as lane closures for the specified period and locations of the lane involved To specify such lane closure or blockage DRACULA Manual v2 4 92 Example Section 6 Simulation Settings specify the lane as a reserved lane and reserve it to DUMMY vehicles see Section 5 4 for the specification of reserved lanes DRACULA does not model irregular incidents such as accidents whose location duration and frequency of occurrence can not be pre defined 6 2 6 Yellow Box Junction To specify a junction as a yellow box simply add a letter Y or y immediately after the node type specification in network coding e g at column 16 of Node Data record of a SATURN network 6 2 7 Motorway Merge The acceleration lane on a motorway merging section can be coded as a dead end lane no turning movements can be made from t
99. hanging models consider the individual driver s intention and ability to change lanes An intention to change lanes will reflect the advantage to be gained eg an increase in speed or an avoidance of delay or the need to do so eg in order to comply with a traffic regulation to avoid an incident in the current lane or to prepare for a turning movement The driver s intention to change lanes may be modelled differently depending on whether the lane change is discretionary or mandatory The intention to make a discretionary lane change may be triggered when the time advantage to be gained by changing lanes exceeds some critical value whereas the intention to make a mandatory lane change will generally be triggered by rules DRACULA divides drivers lane changing desires into one of five types when drivers have to or want to change lane in order to a reach a bus stop on the link b avoid a restricted use lane or incident c make their turn from the next junction d move into a lane reserved for their type or e gain speed by overtaking a slower moving vehicle The first three types are mandatory 1 e the lane changing has to be carried out by a certain position on the current link the other two types are discretionary Whether a discretionary lane change can be carried out depends on the actual traffic conditions For example a vehicle would only change lane to gain speed if the speed offered by the adjacent lane
100. he blue arrow in the picture below Some of them are just wide enough for allow two cars to cross but not a large vehicle with another vehicle Example Fig 4 1 A network model of a narrow bridge over the area modelled DRACULA Manual v2 4 54 Section 4 Basic Highway Traffic Network and Demand There are two types of traffic operations on such narrow roads On some of these narrow roads or sections of a road especially the busier ones traffic signs are put up to indicate which direction of traffic should give priority to vehicles from opposite direction sign in the picture a below and signs to indicate to the vehicles who have priority over the oncoming traffic picture b a Give priority b Have priority to vehicles from over vehicles from opposite direction opposite direction Fig 4 2 Traffic signs used in the UK The second type of operation applies to those quieter less traffic roads Here the roads may be wide enough for two cars but not enough for a large vehicle and another vehicle Traffic operation is this case in terms of who has priority is left to the drivers themselves Two modelling features are introduced to model these narrow roads and access to the road First the narrow road or section of a road should be modelled as a separate link a narrow link The nodes at each end of the narrow link should be coded as of type JTY PE 8 see Section 4 2 1 2 If accessing to the narrow road or br
101. he DRACULA dynamic network microsimulation model In Kitamura R Kuwahara M Eds Simulation Approaches in Transportation Analysis Recent Advances and Challenges Springer pp23 56 ISBNO 387 24108 6 Liu R Van Vliet D and Watling D P 1995 DRACULA Dynamic route assignment combining user learning and microsimulation PTRC Vol E 1995 Liu R van Vliet D and Watling D 2005 Microsimulation models incorporating both demand and supply dynamics Transportation Research 40A 125 150 Liu R Clark S D Montgomery F O and Watling D P 1999a Microscopic modelling of traffic management measures for guided bus operation In Selected Proceedings of 8 World Conference on Transport Research Vol 2 eds Meersman H Van de Voore E amp Winkelmans W Elsevier 1999 Liu R and Tate J E 2004 Network effects of intelligent speed adaptation systems Transportation 31 3 297 325 May A D and Milne D S 2000 Effects of alternative road pricing systems on network performance Transportation Research A 34 407 436 Mauro V 1991 Evaluation of dynamic network control simulation results using NEMIS urban microsimulator Transportation Research Board 70 annual meeting Washington DC DRACULA Manual v2 4 123 REFERENCES McDonald M Hounsell N B amp Kimber R M Geometric delay at non signalised ntersections TRRL SR 10 1984 Milne D S 1997 Modelling the effects of urban road user charging
102. he same network and demand inputs and the same simulation control options the results from two or more runs with different random number seeds may differ To get some confidence in the simulation predictions it is a good practice to run the simulation several times with different random seeds NSEED or NSEED2 and to look at the averages and variances of any result from the number of simulations At the moment there is no automatic procedure within the supply model to help the user to do so The users need to copy the output files from each simulation run either to a different set of file names or to a separate directory All DRACSIM output files are in ASCII text format which can be loaded into standard database such as EXCEL from which statistical measurements such as averages and standard deviations can be calculated The number of simulation runs depends on the level of accuracy required It should consider the level of variance introduced to the model such as variance used in describing the distribution of vehicle characteristics when considering the level of accuracy required Typically 10 simulation runs have been found to be acceptable DRACULA Manual v2 4 118 Section 11 Links with DRACULA Day to Day Model 10 SPECIAL OPTIONS AND FEATURES 10 1 GIS Data Files To enhance background presentation of the study area it 1s possible to include in the network plots geographical features such as rivers railway lines churches and parks
103. his lane as part of a motorway link The length of this link being that of the acceleration section The one ramp traffic entering the motorway can be coded with a turn priority marker S see Section 4 2 1 5 6 2 8 Flared Approaches There are two ways to model a flared approach One method is to insert a dummy node before the flared section and model the flared section as a separate link The DRACULA recommended method is to model the whole link with the number of lanes on the flared section and model the first part of the lane before the flared section as blocked see Section 5 2 3 Example a one lane link of 100m length from node 1111 1112 flares into two lanes on approaching its downstream node 1112 The flared section is 20m and is flared to the kerbside Model solution e code the link with two lanes in the network description file net NET code the first 80m of the kerb lane as a reserved lane which is reserved for DUMMY type vehicles Hint code reserved lane in net BUS file record amp BUS LANE Run DRACPREP and DRACSIM You can change the reserved lane colour using menu Display colour DRACULA Manual v2 4 93 Section 6 Simulation Settings DRACULA Manual v2 4 94 Section 7 Emission and Fuel Consumption Models 7 EMISSION AND FUEL CONSUMPTION MODELS It is well known that vehicles produce more harmful emissions when operating in acceleration and deceleration modes than in cruising mode The
104. ht to be best represented as a separate link in a network Code the guideway links in the network data file under card 1 5 5 Time Dependant Passenger Demand Similar to the model of time dependant traffic demand with fixed route assignment ie Section 4 3 1 3 the passenger flows may also be modelled as time dependant step functions The average passenger flows arriving at each bus stop and its random variation are specified as above see Section 5 2 2 DRACULA Manual v2 4 76 Section 5 Modelling Public Transport By using the following parameters to be specified in PAR file a time dependant passenger demand profile can be described Note that this is a global profile which applies to passenger flows to all bus stops Table 5 3 Specification for time dependant passenger demand Parameter Defaults Description P NSTEP l Number of time steps Maximum 10 steps P_TSTEP 1 DR Duration minutes for the first time period P TSTEP t Duration minutes for time period t t 2 3 10 P GONZO 1 Passenger demand factor for the first time period P GONZO t E Passenger demand factors for time period t 2 3 10 DRACULA Manual v2 4 TI Section 5 Modelling Public Transport DRACULA Manual v2 4 78 Section 6 Simulation Settings 6 SIMULATION SETTINGS The traffic model in DRACULA 1s based on a discrete time micro simulation of individual vehicles space time movements with an time increment of 1 second The simu
105. ic Network and Demand affecting 1t simulation SATURN treats such coding problem simply as a WARNING As described later node coordinates determines some of the basic parameters in DRACULA which affect the results of the simulation Hence DRACULA requires coordinates of all nodes be assigned See Section 4 2 1 6 for DRACULA interpretation of junction layout and entry to stopline link distance 4 2 2 6 Roundabouts All SATURN roundabouts are assumed to operate whereby entry to the roundabout is controlled by Give Way markings and priority must be given to traffic approaching from the right or left for right hand driving Roundabouts which are fully or partially operated by traffic signal controls at the entries must be represented in SATURN as a series of one way links Section 4 2 3 describes a DRACULA model of roundabouts which can take into account both Give way and signal controlled roundabouts A roundabout in a SATURN network is specified in terms of the circulation time and circulating capacity whilst DRACULA requires the physical diameter of and the number of lanes on the roundabout The SATURN network data in node records used by DRACULA to represent a roundabout are RSAT Maximum roundabout capacity in pcu hr JCIR Roundabout circulating time seconds By default the number of lanes on the roundabout NRBL is estimated from the maximum roundabout capacity RSAT assuming 1800 veh hr lane and the maximum number
106. idge is controlled by give way sign sign a then the entry arm to the bridge should be coded with a G turn priority marker TPM Section 4 2 1 5 The opposite with high priority would have TPM BLANK If the narrow road is not controlled with a give way sign then the entry arms from both direction should have TPM BLANK DRACULA Manual v2 4 55 Section 4 Basic Highway Traffic Network and Demand In the example of Sutton Bridge the section of road over the bridge 1s modelling as link from 103 104 for the east bound traffic and link 104 to 103 for west bound traffic A give way to oncoming vehicles sign is put up for the east bound traffic indicated in the network plot by a solid stop line downstream of link 102 to 203 The node and link coding in card 11111 of SATURN format for the bridge is as follows and node the turn 102 103 140 has a TPM of G Example E ES RAW DL Fig 4 3 An example coding of a narrow bridge fon A Note 1 Such narrow links can only be connected to one other link e g junction of type 8 could only have two arms e g NIN 2 Section 4 2 1 2 11 There could be only one lane on such links The second feature introduced to the model is one used to describe the characteristics of the narrow bridge or link itself and to model the traffic operation over the bridge A narrow link is characterised by its width measured as a fraction of the normal lane width The
107. ile OUT for each link and for the whole network at the end of current time period Performance Measures Queuing delay veh hrs Cruising time veh hrs Total travel time veh hrs Total travel distance veh kms Average speed km hr Performed flow veh km hr Average flow veh hr Fuel consumption litre Total CO emission Kg Total NOx emission Kg Total HC emission Kg Supply Measures Total trips demanded veh Total travel time veh hrs Total travel distance veh kms Average flow veh hr Average speed km hr Average cost pence DRACULA Manual v2 4 116 Section 9 Model Outputs and Evaluation 9 5 Exit Measures These are data collected when vehicles exit a link or the network They do not include vehicles still travelling on the link or in the network When a vehicle exits a link or the network its travel time and distance over the link and throughout the journey are recorded Travel time is averaged over the number of vehicles recorded An average speed is derived from the average travel time and distance The measures are disaggregated by vehicle types The program outputs for each link and for each vehicle type at the end of the current time period e Link free flow travel time second Number of vehicles exiting the link during the current time interval Average travel time in the current time interval second Standard deviation of the travel time second Maximum travel time in the current time in
108. imum deceleration desired speed relative to the mean speed on any individual link and a gap acceptance parameter These characteristics are randomly sampled from normal distributions representative of that type of vehicle See 6 1 3 Public transport vehicles are represented with additional information such as service number service frequency bus stops and average passenger flows at each bus stop etc See 5 Vehicle movements in a network are determined by its desired movement such as desired speed lane choice response to traffic regulations and interactions with neighboring vehicles The simulation maintains a linked list of vehicles in each lane and moves individual vehicles according to a car following model and a lane changing model and their response to traffic controls at intersections The car following model calculates a vehicle s acceleration in response to its desired speed and the relative speed and distance of the preceding vehicle Depending on the magnitude of the relative distance a vehicle is classified into one of three regimes free moving following or close following The lane changing model contains three steps 1 obtain the lane changing desires and define the type of changing 2 select the target lane and 3 change lane if all gaps are acceptable The model divides drivers lane changing desires into one of six types when drivers have to or want to change lane in order to a reach a bus stop o
109. ing vehicle in front and anticipates to catch up within time TRUB a desire for overtaking will be triggered The driver will then look for overtaking opportunities If there is enough space NLENGTH in front of the vehicle for merging back after the overtaking and the gap in the opposing traffic is sufficiently large OGAP and it is within a visibility range SRUB the driver will begin the overtaking manoeuvre at a speed faster and acceleration higher than his her normal speed and acceleration by a factor of FVMAX and FAMAX respectively in order to complete the overtaking quickly The parameters specifying the overtaking behaviour are listed in Table 8 2 they can be set in PAR Some of these default values are from TRARR model those with others are simply guess work which need sensitivity tests To specify the links without double white line markings enter the relevant links in the additional network data file ADD between amp RURAL and 99999 delimiters If a link is not specified by default it prohibits overtaking using the other side of the road DATA NAME DESCRIPTION l ANODE Upstream node number of the link 2 BNODE Downstream node number of the link 3 VTYPE Type s of vehicles permitted for overtaking DRACULA Manual v2 4 108 Section 8 Advanced Simulation Applications The following example suggests that overtaking by all types of vehicles 1s permitted from link 101 109 whilst only CAR type is permi
110. l and stage extension or recall respond to the first bus detected In all cases the constant intergreens must be maintained for safety reasons DRACULA Manual v2 4 103 Section 8 Advanced Simulation Applications 8 2 3 Payback If an extension to the bus link green stage is triggered in one cycle the extension time is removed from the bus link green stage in the following cycle subject to a minimum safety green This removed time is subsequently given back to the other link stage In the case of an early recall the amount of recall time is given back to the bus red stage in the following cycle This procedure is named payback It is designed to maintain signal offsets and the long term degree of saturation on all links 8 2 4 Cooling Off Period After an extension or recall in one cycle and a payback in the following cycle there will be a recovery period during which no stage change is permitted This period is named the cooling off period and is intended to let the system settle down The period is modelled as a multiple of the signal cycle via parameter NCOOLING CYCLE 8 2 5 Plan Selection Signal Control This signal control strategy was supplied by Stuart Dagleish of York City Council and implemented in DRACULA to help design a quality bus corridor scheme in York The traffic signals are varied between two alternative fixed plans in response to detection of a selected type of vehicles in practice buses The alternative signa
111. l plan is called in at the detection of a bus upstream and cancelled at the detection of the bus downstream While the signal is still on the alternative plan further detection will reinforce the use of the alternative signal plan until they have all passed the downstream detector A cooling off period is enforced after a round of signal changes to ensure the system to settle down well The initial fixed plan signal is coded in the network description file The alternative fixed plan signal is coded in the detector data file There need a pair of detectors one upstream and one downstream of the traffic signal They are of detector type T and R for the upstream detector and the downstream detector respectively The detectors and information on the alternative signal plans is required in the data file in a format similar to that in DRACULA free format network coding See Section 8 1 for coding of detectors and alternative signal plans DRACULA Manual v2 4 104 The control strategy Data inputs in DRACULA Section 8 Advanced Simulation Applications 8 3 Speed Advice and Intelligent Speed Adaptation A mandatory speed control system or an advisory speed control system with known compliance rate can be modelled m DRACULA The model was developed initially to represent the Intelligent Speed Adaptation ISA or External Vehicle Speed Control EVSC systems Liu amp Tate 2004 ISA 1s a technical device fitted to the vehicles enabling the
112. lated NOx emission in kg Accumulated HC emission in kg Accumulated fuel consumption in litres DRACULA Manual v2 4 117 Section 9 Model Outputs and Evaluation 9 7 Definition of Queue Length The queue length of a lane is the distance from the stopline upstream to the end of queue in meters The queue length of a link is the longest queue length of all the lanes in the link The end of queue of a lane is defined as the back of the first stationary vehicle from upstream start of the lane Stationary is defined as speed less than 0 5m s The queue length for each link is recorded at every second of the simulation At the end of each measuring time period the program outputs the time averages of link queue length variance and the maximum queue length during the measuring time period in the following format Sample time period in sec incremented by sec Time average queue length in metre in the current time interval Standard deviation of the queue length Maximum queue length in the current time interval Sample time period in sec from the start of simulation Accumulated average queue length in the current time interval Standard deviation of the accumulated queue length Accumulated maximum queue length sec 9 8 Multiple Simulation Runs DRACULA is a stochastic simulation model there are many random processes used occurring during the simulation e g vehicles random parameters and drivers random behaviour Given t
113. lation is based on individual vehicles characteristics their car following lane changing and gap acceptance behaviour and traffic controls in the network Simulation results are affected by the input vehicles characters on parameters describing their driving behaviours and on conditions and controls in the network This section describes the settings that could affect the simulation These settings are divided into global ones which affect the whole network and throughout the whole simulation time periods and local settings which affect only parts of the network and during parts of the time 6 1 Global Settings 6 1 1 Simulation Periods Three time periods were simulated in DRACULA these are schematically depicted in Fig 6 1 Demand Factor F2 Demand period Cooling period Time aih tend ay Mr T2 Toan 1 T2t T3 Fig 6 1 Simulation time periods and representation of temporal demand profile The demand period is the main simulation period specified via parameter PERIOD in DRACULA network coding Section 4 2 1 and LTP from a SATURN network Section 4 2 2 The demand period is typically one hour representing the peak period There could be a number of sub periods within the main study period to represent the temporal distribution of demand Section 4 3 1 2 The demand levels for each time period Fi F2 etc in Fig 6 1 are represented as a proportion of some nominal d
114. mation 8 2 2 Signal Extension and Recall When a bus is detected at time to and predicted to arrive at the stopline at time t one of two actions may be performed Extension which extends the bus green period in order to allow the bus to exit Recall which terminates the bus red stage earlier in order to reduce the bus waiting time Figure 8 1 shows schematically the signal priority in a space time diagram Three bus trajectories from the detector to the stopline are drawn in dashed lines Distance IG Green Amber Red IG Green Bus link signal stopline detector time Figure 8 1 Space time representation of bus signal priority If a bus is predicted to arrive at the stopline just after the start of the red signal the bus green stage and the following amber will be extended by just enough time to allow the bus to exit The amount extended depends on the predicted bus arrival time subject to a user defined maximum parameter EXTENSION and to a minimum safety green for the subsequent stages affected If a bus is predicted to arrive during the red but an extension is not appropriate then the duration of the bus red stage may be reduced by a constant amount defined by parameter RECALL The stage length of other stages remain un changed the length of the current cycle is therefore changed Otherwise the signals will not be changed case A in Figure 8 1 Only one signal change extension or recall is permitted per cycle
115. ments within which the following tasks are performed a Initialisation Set within day simulation clock t 0 b Vehicle Generation Generate new entry vehicles at their preferred departure time and place them on their entrance links c Vehicle Movement Loop through all vehicles in the network and for each of them 1 check if the vehicle wants to change lane if so whether the gaps are acceptable ii update the vehicle s speed and acceleration and advance it to its new position At the end of the link either remove the vehicle from the network if it arrives at its destination or pass it to its next link en route iii Calculate vehicle emissions and fuel consumption and record traffic performance measures DRACULA Manual v2 4 2 Simulation Loop Network Representation Section 1 Introduction d Animation Update Update the graphical display if required e Traffic Control Update For each signalised junction update the stage change over clock according to desired signal plans fixed plans or responsive Check if the any incident is to start or to finish f Time Update If all drivers have finished their journey terminate the day otherwise increment the simulation clock and return to step b The simulation loop is depicted in Figure 1 1 Initialisation Vehicle generation Vehicle movement Traffic control update Time update Repeat until all drivers have reached their destinations
116. meters for a given link use the additional network inputs in text file ADD Insert a section in the file beginning with amp GAPS and ending with 99999 Enter the links and the gap values in between the start and end delimiters following the format below DATA NAME DESCRIPTION ee LINK GAP DATA ANODE Upstream node BNODE Downstream node GAP average gap value seconds of the link MIN GAP minimum gaps seconds of the link GAP TSTART time sec since waiting before reducing gaps GAP TEND time sec since waiting before taking MIN GAP Nn B WN mm The example below suggests that vehicles approaching node 21 along link 31 21 take an average gap of 5 seconds during the first 60 seconds of waiting time then gradually reducing their acceptable gap to the minimum gap of 2 seconds at after having waited 120 seconds Vehicles along link 55 21 have an average gap of 4 8 sec minimun gap 1 2 sec and two waiting times are 80 and 160 sec respectively net ADD amp GAP 31 21 5 2 60 120 Example 5521 4 8 1 2 80160 cae ees 99999 amp STOP 5 20 5 5 99999 amp RZONE 5 20 85 0 99999 amp CIRC 1115 20 DRACULA Manual v2 4 90 Different gap acceptance values Section 6 Simulation Settings 1116 30 1 20 2 4 99999 The rules for using various level of gap values are 1 by default all approaches links use the default global gap values e g GAP 3 5 seconds MIN GAP 1 sec 2 If gap values are specified m the
117. n the link b avoid a restricted use lane or incident c make their turn from the next junction d move into a lane reserved for their type e gain speed by overtaking a slower moving vehicle by changing to a lane in the same direction or f overtake a slower moving vehicle by changing to a lane in the opposite direction of traffic DRACULA Manual v2 4 4 Vehicle Characteristics Vehicle Simulation Simulation Outputs Section 1 Introduction As a measure of network performance the simulation outputs by default network link and route specific measures such as total vehicle hour total vehicle km average travel time speed queue length fuel consumption and pollutant emission over regular time periods The length of the report time period can be defined by the user At the user s request the program may also output individual vehicles second by second locations and speeds to provide space time trajectories of the vehicles A graphical animation of the vehicles movements can also be shown in parallel with the simulation giving the user a direct view of the traffic conditions on the network See 49 A new program called SPATULA 9 2 has been written to convert the link based simulation results into a SATURN style data file which can then be input to a SATURN program P1X for display and for comparison with SATURN simulated results DRACULA Manual v2 4 5 Section 2 System Requirement and Model Architecture
118. nced Simulation Applications 8 ADVANCED SIMULATION APPLICATIONS 8 1 Selective Vehicle Detectors Detectors are placed on the road as line sources on a lane If at time to 1 seconds the front bumper of a vehicle is before a detector and at the following second to its front bumper is at or has just passed the detector regardless whether the rear end of the vehicle has or has not passed the detector then a vehicle detection is triggered The actual detection time in 0 1 second is interpreted from the speed and location of the vehicle at time t With such a definition a vehicle stopped on top of a detector will not trigger more than one detection The input data consists of e detector identification e the lane where the detector is placed e location of the detector from the stopline DRACPREP scans for detector inputs in the DET file automatically It tries to locate a DET or file name specified via parameter DETFIL in the PAR file The detector file contains two parts 1 detector data and 2 inputs of an alternative signal plan The second part of the data is linked to a plan selection responsive signal control which is to be explained in Section 8 3 5 There are two sections of records l Section 1 Detection description Preceded by a line containing a string amp DETECTOR and terminated by a line containing 99999 One record per detector PA Section 2 Description of alternative signals Prec
119. network performance measures and the supply costs of trips 6 1 2 Vehicle Types Currently there are seven types of vehicles defined in DRACULA Each type of vehicle 1s represented by a name and a unique number The names and the numbers of the seven types of vehicles are listed 1n the table below Table 6 2 Definition of vehicle types DUMMY 0 Dummy vehicles Bus type Bustypel S TAXI J8 Tai O The names are given according to their default types for example type TAXT of number 8 represents by default taxi However there is no reason why users cannot re define the types according to their needs as long as you specify the right characteristics for the right type see next sub section For example you could user TAXI to represent high occupancy vehicles or use it for trams as long as you give the appropriate specification of the characters for your chosen types of vehicles Take another example if you need to have three different categories of buses for example ordinary buses express buses and P amp R buses you could borrow of the other types that you don t need and re define the characters accordingly The only limitation is that you can not have more than six vehicle types type DUMMY is reserved for some special use and cannot be replaced Development is underway to incorporate more vehicle types or user classes into the model DRACULA Manual v2 4 81 Section 6 Simulation Settings 6
120. nk LANES Number of entry lanes for this link 0 if the link is one way from NODE to ANODE 3 SPEED Link free flow speed kph 4 IDIST Distance meters measured from the centre of NODE to the centre of ANODE See 4 2 2 Data for the first turn from this link clockwise from left for left hand driving and anti clockwise from right for right hand driving condition 5 LANEI First lane used by this turn counted from the nearside kerbside kerbside lane 1 0 if the turn is banned 6 TPM Turn priority marker for this turn This is to be entered immediately following LANEI there should be no space between LANEI and TPM See 4 2 1 5 7 LANE2 Last lane used by this turn 0 if the turn 1s banned Data for the second turn from this link Same entries as for the first turn Data for the last turn from this link The maximum number of turns being possible or banned equals to the number of arms to the junction Same entries as for the first turn rer RECORD TYPE 3 SIGNAL DATA One record for each of the NSTAGE stages l STAGL Duration of stage seconds 2 INTG Duration of following inter green seconds 3 GNA 1 The entry node for the first turn 4 GNC 1 The exit node for the first turn 5 GNA 2 The entry node for the second turn 6 GNC 2 The exit node for the second turn etc etc with all permitted turns coded on the same line DRACULA Manual v2 4 33 Section 4 Basic Highway Traf
121. normally give way turn G turn which makes alternative turns with the priority stream into the junction Section 4 2 1 5 SATURN distinguishes hooked or not hooked opposite pairs of right turning G or X movements In DRACULA such pair of movements is always modelled as not hooked Hence DRACULA discards SATURN turn modifier D and E The only SATURN turn modifier TPMod used by DRACULA is C for a clear or reserved exit lane for G or X movements This is modelled in the same way as the C turn in Section 4 2 1 5 As with the DRACULA format coding of traffic signals there could be at most two links in any one stage There are however no limit on the number of turns that can be coded in each stage as long as the turns are from the two links 4 2 2 4 Simulation Zone Data Currently the SATURN simulation zone data are used in by DRACULA in two ways 1 to aggregate simulation statistics by origin destination pairs and 11 to be used in conjunction with the day to day demand model Section 10 The zone data are not affecting the simulation directly Hence it is not crucial to have zones coded or embedded in the main NET file 4 2 2 5 Node Coordinates and Link Length If the coordinates of a node are not assigned SATURN tries to extrapolate interpolate its coordinates from those of its neighbours Since the node coordinates are used purly for display purposes in SATURN not DRACULA Manual v2 4 46 Section 4 Basic Highway Traff
122. of lanes on any entry lane NLentry NRBL max RSAT 1800 NL 4 1a Alternatively one can make NRBL depend only on RSAT as NRBL RSAT 1800 4 1b To choose equation 1b set a parameter QRBLANE F in PAR file DRACULA makes an assumption about the free flow circulating speed and then estimates the inscribed diameter of the roundabout from JCIR HCID max R JCIR xV circ 2m NRBLxW p 4 2 island where HCID half of the inscribed diameter 1 e radius of outer boundary of a roundabout Ristand a user defined minimum radius of the central island metre Wiane the width ofa lane which is fixed to 4 metres and DRACULA Manual v2 4 47 Section 4 Basic Highway Traffic Network and Demand Voir the free flow circulating speed Default 27 km hr Variable Vir can be specified via a control parameter CIRC SPEED in the parameter file net PAR The default circulating speed of 27 km hr is taken from a study by McDonald et al 1984 This is a global variable which applies to all roundabouts in a network To make a roundabout specific circulating speed see description in Section 6 2 If the diameter and the number of lanes on any of the roundabouts in the network are known one can over write the default estimations with the actual data by manually altering the values of circulating time JCIR and the maximum roundabout capacity RSAT to match those observed Alternatively one can define HCID Veic Risiang
123. of the outputs are explained at the top of each file and or of each section e lt net gt OUT contains statistical measures on traffic congestion defined by a performance and a supply measure of congestion for link route and network at regular time periods e lt net gt SPA contains link based performance measures same as in lt net gt OUT at regular time periods in a format to be used by the program SPATULA Section 9 2 for SATURN e lt net gt REB contains statistics collected at link and network exits on queue length flow travel time and travel speed and variability of these measures Also contained are speed distribution profiles e lt net gt POL Statistics on pollutant emission and fuel consumption measures for each link and for the whole network at regular time periods e lt net gt LTT Reports on individual vehicles route departure time arrival time and link by link travel time e lt net gt PTT Same as lt netname gt Itt but for buses only They are also reported in lt netname gt Itt e lt net gt PSN Records of individual buses dwell time and passenger delay at bus stops e lt net gt TXD Individual vehicles speed and time when passing detectors e lt net gt TXS General text reports Also reports on each rural overtaking events overtaking made from the opposite road space e lt net gt TRJ when specified the program can also output second b
124. ol If a link is not specified it will be a link that is not under the speed control and is flat has zero gradient There could be more than one speed limit specified for one link The format of the inputs are DATA NAME DESCRIPTION l ANODE Upstream node number of the link 2 BNODE Downstream node number of the link 3 NSPEED Number of speed limit sections on link 4 START 1 Start position of the first speed limit section metre relative to link entry 5 END 1 End position of the first speed limit section metre relative to link entry 6 SPEED 1 Speed limit kph for the section etc etc for each speed limit section 7 START n Start position of the last speed limit section metre relative to link entry n NSPEED 8 END n End position of the last speed limit section metre relative to link entry n NSPEED 9 SPEED n Speed limit kph for the section Use control parameter PISA in PAR to specify the ISA system penetration rate Defaults Percentage vehicles fleet under ISA control If PISA 10 then 10 penetration DRACULA Manual v2 4 106 Section 8 Advanced Simulation Applications 8 4 Modelling Dynamic Road Pricing Systems Two dynamic road pricing systems have been implemented in DRACULA they are time based pricing and the Congestion Metering system the latter was first proposed for Cambridge by Oldridge 1990 and tested in an EU project ADEPT Clark et al 1993 The underlying concept of the Con
125. on traffic follows RLANES in each section The example given in 4 2 3 2 shows a roundabout operating as 4 lanes on half of the circulating carriageway and 5 lanes on the other half Careful observation of the animation would reveal that the inner most lane on the 4 lane half of the carriageway is never being used by traffic For large roundabouts there are often separation between an entry and an exit of an arm RSHIFT is to represent partly such physical separation and partly the entry angle of a roundabout approach The NIN number of RBNODE needs to be entered in strictly clockwise order or anti clockwise for driving on right and in the same order as the ANODE being entered in the Node Record Section 4 2 1 2 Lane marking is given such turning left is 2 right 4 and straight 8 A shared lane is simply to add the two turning numbers together 1 e a RLMARK 10 indicates a shared left and straight ahead turnings 50 Section 4 Basic Highway Traffic Network and Demand 4 2 3 2 An Example The network description of the Stanwell Moor Roundabout which 1s partially signalised and with spiral marking 1s given below See the additional coding for B node 105 Example r EC Sie M25 Stanwell Moor Roundabout June 1998 with Indirect Signals on A3044 North 11111 1101 1 0 101 2 70 200 101 2 4 1101 4 70 200 1800 14 105 2 70 200 1800 1 2 1102 1 0 102 2 70 200 102 2 o 1 28 50 1102 4 70 200 1800 14 105 2 70 200
126. on 3 6 windows based front ends or via batch mode commands under Command Prompt Section Ap 3 1 DRACWIN Front End The DRACPREP and DRACSIM can be accessed via a Windows front end for DRACULA from a program called DRACWIN EXE Double click the icon DRACWIN on the desktop or DRACWIN EXE the interface DRACULA WINDOWS CONTROL pops up to load the front end GITS DRACULA WINDOWS CONTROL Programs Folder Module Run Command Line Run Demos Tools Help URAGULA There are six menus on the top main menus Programs Folder Module Run Command Line Run Demos Tools Help To change the path of executive files e g DRACSIM exe DRACPREP exe and SPATULA BAT To execute the programs DRACPREP DRACSIM SPATULA To run the executive programs with DOS Command To display DRACULA simulation runs on demonstration networks To access DOS Prompt Version Information DRACULA Manual v2 4 13 Section 3 Executing the Programs 3 2 Running the SATPIG Program SATPIG is a SATURN program to produce a text file of origin destination route flows from a SATURN assignment It was written in particular to facilitate interface with DRACULA its output route flows are written to the trp format required by DRACULA To start SATPIG select from main menu Module Run SATPIG DRACULA WINDOWS CONTROL Programs Folder ahr Command Line Run Demos Tools Help SAT PIs DRACPREP DRACSIM SPA
127. operation of the traffic over the narrow bridge is characterised by the types of vehicles to give way and a sight distance over the other end of the bridge The data should be coded in the additional network input data file ADD and within the headings amp BRIDGE and 99999 as follows DATA NAME DESCRIPTION l ANODE Upstream node number of the link Note 1 2 BNODE Downstream node number 3 VTYPE Vehicle types to give way See Note 1 4 FWIDTH Fraction of a normal lane width Note 111 5 SIGHT Sight distance metre Note iv Note 1 A narrow bridge is still a two way road coded in the main network description as two uni directional links Both these links should be coded here in ADD 11 A give way to oncoming vehicles sign normally indicates give way to all types of vehicles Here VTYPE should be coded as 63 see Section 6 1 2 for definition of vehicle types Where a road is DRACULA Manual v2 4 56 Section 4 Basic Highway Traffic Network and Demand only narrow for a large vehicle plus another vehicle VTYPE can be used to indicate the actual types of large vehicles which are of problems In the Example 2 below the road is wide enough for two cars but are too narrow if there is a LGV or HGV vehicle present VTYPE 48 is the sum of type 16 vehicles LGV and type 32 vehicles HGV 111 The normal lane width is modelled as a fixed value at 4 metre s wide Given that this is a narrow link hence
128. or the program Table 2 4 Input data files for DRACSIM Data Type Details of Input Data File name and extension Basic inputs Network supply description SUP mandatory List of routes and aggregated route flows Selective inputs discretionary Individual trip departure time and vehicle characteristics Bitmap background file and associated coordinates Section 10 2 Global input Specification of vehicle characteristics VEH TAB discretionary Section 6 1 3 Emission and fuel consumption factors POL TAB Section 7 The basic and selective inputs are network dependent whilst the global inputs are not DRACULA Manual v2 4 10 Section 2 System Requirement and Model Architecture Depending on the choices provided by the user in the parameter control file different levels of details are output ranging from the most detailed second by second vehicle trajectories to aggregated link route and network measures Format of the output files are described in Section 9 Table 2 5 Outputs from DRACSIM Data type Details of Output Files Default Text message reporting the status of TXS simulation Summary statistics on travel OUT performance measures Network and link performance SPA measures for use with SPATULA LTT T RJ Summary statistics on pollution POL measures Individual vehicles link travel time Optional Individual buses link travel time Summary of passenger delays
129. or vehicles that are in idling and decelerating modes and vary as a function of speed for cruising vehicles For vehicles that are in acceleration mode the fuel consumption factor f is a function of both speed v and acceleration a of the vehicle f c c xaxv 7 1 where Cy and c are two constants given in Ferreira 1982 The models are developed to be flexible so that new values in terms of new emission rates for the existing pollutants and for new pollutants can be incorporated as they become available The new factors can be fed into DRACSIM via a look up table POL TAB so long as the relationship between traffic condition driving mode and emission remains the same The following example file contains the default factors used in DRACSIM DRACULA Manual v2 4 95 Section 7 Emission and Fuel Consumption Models POL TAB Pollution Table Emission rates g sec for 7 different types HGV and pollutants amp POL RATE CAR CO d 10 0 0 06 0 06 0 377 0 072 NOx 7 MO gO 200 0 0006 0 0006 0 0017 0 0008 0 01 0 0005 HC 7 10 0 2040 0 0063 0 0078 0 0083 0 0067 0 02 0 0067 FUEL 12 10 0 20 0 30 0 40 0 T200 0 26 0 29 0 34 0 40 DOS 0 42 0 26 Bis e amp ENDPOL BUS CO T 10 0 0 06 0 06 0 377 0 072 NOx 7 io 200 0 0006 0 0006 0 0017 0 0008 0 01 0 0005 HC 7 10 0 20 0 0 0063 0 0078 0 0083 0 0067 0 02 0 0067 FUEL 12 10 0 20 0 30 0 40 0 Ae een 0 26 0 29 0 34 0 40 CRC rere 0 42 0 26 O25 am
130. out capacity Section 4 2 2 3 QSATNET T if to take SATURN format simulation network F to use the free format network coding as described in Section 4 2 2 QSATPIG F if to take routes and route flows from input file TRP of format else 1f T take them from LPG file Sections 4 3 1 4 3 3 QVEHPOOL Set T to enable DRACPRPEP to generate individual vehicles list in VEH file Section 4 3 2 TRPNAM t Name of input TRP file for time period t Section 2 2 3 TSTEP t l Length min of time period t Sections 4 3 1 2 4 3 3 3 XYUNIT 1 0 Factor converting one unit of node zone coordinate to metres If XYUNIT 10 then 1 unit of coordinate 10m Same as used in SATURN so can also be specified in the network description file NET Section 4 2 1 DRACULA Manual v2 4 126 APPENDIXES APPENDIX B List of Simulation Control Parameters DRACSIM Defaults Definition AMBER PERIOD Amber time period second ARMCOLOUR 4 Colour of displayed clock arms or Default colour RED COL ARM Section 3 7 2 2 9 COL FACE Ditto for the clock face Default colour BLACK BKCOLOUR Ditto for background display or Default colour LIGHTBLUE COL BKGROUND Default colour GREY COL LANE 15 Ditto for lane markings Default colour WHITE DM EDM DARKGREY Default colour DARKGREY Ditto for bus lane Default colour LIGHTCYAN CHRGRATE 1 0 Charge rate pence min Section 8 5 CHRGTC 3 0 Charge threshold min Section 8 5 CHRGSEG 10 Charge segm
131. oute choice departure time choice and characteristics for cach vehicles One record per line and for each of the NVEH vehicles in ascending order of departure times DRACULA Manual v2 4 62 Section 4 Basic Highway Traffic Network and Demand DATA NAME DESCRIPTION l NVEH Total number of vehicles to travel PERES RECORD TYPE 2 INDIMIDUAL VEHICLES TEKEE 58 l VEHID Vehicle identification a unique number DEPT Departure time seconds measured from the beginning of the demand period 3 RTID Route identification 1 a number pointing to the i th route specified in file TRP 4 VTYPE Vehicle type see 6 1 2 for definition 5 LEN Length of the vehicle metres 6 SMIN Minimum safety distance headway metres 7 TREAC Reaction time seconds fixed at 1 0 second 8 ANORM Normal acceleration m s 9 AMAX Maximum acceleration m s 10 DNORM Normal deceleration m s 11 DMAX Maximum deceleration m s 12 SPEED Speed factor 13 GAP Gap acceptance factor The vehicle characteristics CDATA entries 4 13 are drawn from distributions which try to match as closely as possible the behaviour of drivers in the study area Details on the distributions used can be found in Section 6 1 3 4 3 3 Route Assignment from SATURN A SATURN program SATPIG can be used to produce a file of origin destination route flows from a SATURN assignment To run SATPIG use a command line such as SATPIG net trips wh
132. overtaking is allowed in opposite road space and normal roads where vehicles are not allowed to move over to the other side of the road in one network 8 5 New type of junction control give way to on coming traffic Associated is a new link type links with just one lane and a narrower than normal lane This is to represent narrow roads such as a narrow bridge where one direction of traffic has to give way to traffic from the opposite direction 4 2 1 and 4 2 5 Model of traffic calming measures 6 2 Removed constraints on how many links can be in any one stages This applies to both the DRACULA format network coding 4 2 1 2 and the loading of SATURN networks The default vehicle type LGV should be called Light goods vehicle instead of Large goods vehicle They include 2 axil transit vans 6 1 2 Allow user specification of the speeds at which vehicles enter the network By default vehicles start with a zero speed at entry The car following model will determine the speed of the vehicle at the next time increment Through the new parameters INIT SPEED and INIT OLD SPEED users can specify the global initial speeds in kph or if one of the two new parameters is set to 1 the local free flow speed of the entry link will be assigned to the vehicle This is coupled with parameter NENTRY to allow user selection of the entry speed of the vehicles Appendix B Add coefficient of variation in passenger arrival dis
133. ow curves These parameters can be measured and have been used by earlier authors to describe the way in which costs of using the network rise as usage increases DRACULA Manual v2 4 115 Section 9 Model Outputs and Evaluation Supply curves reflect the costs experienced by a driver at a given level of demand which will be affected by the journey length and the route taken as well as by the impacts of other demands on the network both at the same time and in earlier time periods Supply curves cannot be readily observed in the way that performance curves can The space time domains used to measure performance curves and supply curves are different as shown in Figure 9 2 The supply curves were generated through a vehicle tracking approach May et al 2000 whereby individual vehicles journey time and distance from origin to destination were used to give a generalised costs The resulting generalised costs were then summarised by vehicles departure times to give supply costs for the time period of interest The supply curves are then represented as generalised costs per trip vs the trips demanded over a given departure time period The performance measures were generated through a time slice approach whereby vehicle trajectories time and distance within a specific time period were used to get the generalised cost for the period There was no distinction as to the departure time of individual vehicles The program outputs in f
134. p ENDPOL amp END Ze BO a0 O RR O ZL Owes ZO SO OO HERS PMES DRACULA Manual v2 4 CO NOX HC 4 0 Oe LeS SOROS Oh Oe LA 30 0 40 30 0 40 40 Oe Z oS DO sd a dp 30 0 AGG 30 0 40 and Fuel comsumption rates ml sec of vehicles DUMMY CAR BUS GBUS FUEL O 13000 60 000 2050 Oo S000 26040 4040 Ds 00Z2 DVD AZ 00050 0a 0056 O 2040 6040 00 gUll 00097 OC lly 0 0136 SU ei SOO AQ SO 0 GOLO dO OO DIDO Du AS Ooo Duda Oe 2 Lodo AS Lobi ad O 2008 600 FO O SOmwb 260 00 AUD V40022 0 0042 04 0050 020058 O 23020 GOO 7040 QUIZ Oe 0097 Os OLLO OOL SU Oi GOO FOG SO 0 GOLO dO OO IVO Oe Ooo CTE One 2s Lodo O Lie aed 96 Ley Example Section 7 Emission and Fuel Consumption Models It is possible to output pollution by individual vehicles In the current version only accumulated emission and fuel consumption measures for each link and for the whole network are stored and output at regular time intervals DRACULA does not model pollution dispersion The modelled emission by link and by time period can be easily linked to dispersion model packages available such as ADMS Tip Due to the many uncertainties in the model inputs the pollution model is better suited for comparing strategies than for producing absolute emission levels DRACULA Manual v2 4 97 Section 7 Emission and Fuel Consumption Models DRACULA Manual v2 4 98 Section 8 Adva
135. p acceptance factor a E a 2 Max acceleration m s ee Nel iad es ae The risk factor here represents a gap acceptance behaviour see Section 6 1 3 for full description A risk factor of 0 5 1s used for the two types of buses to represent the bus drivers lane changing behaviour they tend to take smaller gaps than car drivers to push into the main stream of traffic or to an adjacent lane 5 1 3 Modelling Bus Stops and Bus Dwell Time An ordinary bus stop is a single sign on the road side and a bus stops alongside the sign on the road to pick up or put down passengers thereby blocking the following traffic in that lane A bus layby however provides a space for the bus to pull into at the bus stop and thus allows following traffic to pass A bus layby in the model is represented as a special type of bus stop A bus stop is described by the following data e the bus stop identification e the link it is on e location of the bus stop on the link measured from the entry of the link e type of bus stop stop or layby e length of the stop in case of layby e lateral location on the link kerb side or median DRACULA Manual v2 4 68 Section 5 Modelling Public Transport e average passenger arrival rate passengers hour at the bus stop There can be several bus stops on one link One bus stop can be used by more than one bus service Before entering a link with a bus stop a bus will try to get into a lane which Bus
136. ppropriate lane according to their turning exit and to the lane markings on the approaching arms In this way they would get into the appropriate lane on roundabout and change lane on the roundabout only to make their exits The workings of the roundabout in terms of vehicle joining circulating and leaving the roundabout and lane changing on the roundabout are all worked out internally by the model Therefore when specifying the path route of a trip you only need to give the node number of the roundabout and not the node numbers of each approach For the example in Section 4 2 3 2 a path from zone 5001 to zone 5003 will go through nodes 1101 101 105 103 and 1103 In the traditional way of representing a roundabout as a series of one way links such lane choice behaviour of traffic approaching a roundabout gets lost as all the lanes on the approaching arm permit the same turn There would also be too much unrealistic lane changing on those one way roundabout links 4 2 4 Model for Give way to the Right Junction Control In the Netherlands and part of Belgium a type of junction control is such that one should always give way to all traffic from the right for driving on the right This type of junction is modelled as a type 9 junction in DRACULA see node coding in Section 4 2 1 2 The following example shows the general rules for this type of traffic control The example is for driving on the left though he model can also be applie
137. r Xpcu to DRACULA flow in veh hr Xen as follows Xen X pu PCU 4 3 Hence a SATURN flow of 200 pcu hr with PCU 2 would result a flow of 100 vehicles hr for DRACULA Assuming that there are three user classes car taxi and HGV using the lt net gt road network passenger car unit equivalent for car is indicated as VCPCU 1 1 VCPUC 2 1 3 for large taxi and VCPCU 3 for HGV It is easy to see that the reference number for each user class should match with the level for that user class as defined in the stacked matrix of travel demand An example of PAR is given below Example PARAMETERS pele VCPCU 1 1 VCPCU 2 1 3 Go 5 VCPCU 3 3 END em Q j mf on zag E Bus flows in Buses in SATURN are coded as fixed flows along the specified routes via SATURN 66666 records in the DAT file They are coded in unit of service frequency i e number of buses per hour per service Bus flows are not included in the SATURN demand matrix When SATURN loads the bus flows it first DRACULA Manual v2 4 64 Section 4 Basic Highway Traffic Network and Demand converts them from buses hr into PCUs hr after multiplying by the BUSPCU factor defined in the SATURN network parameters list then assigns them on to the road network As DRACULA simulates individual vehicles as opposed to traffic flow measured in PCUs it uses the SATURN bus frequency data specified in the 66666 records directly Hence there is no n
138. r ISA control Section 8 4 Proportion of total number of trips are made by LGVs Section 6 1 4 Pence per km used to convert distances into generalised costs Section 9 Pence per minute used to convert times into generalised costs Section 9 Proportion of total number of trips are made up by taxis Section 6 1 4 Percentage of vehicles trajectories to be output Example PTRAJ 10 for 10 If T display vehicle length in proportion to the link length and crow fly distance Section 4 2 2 2 DRACULA Manual v2 4 129 APPENDIXES QCHRG F TRUE if to calculate congestion road pricing Section 8 5 QDET FULL If T output full detection report into file TXD QEXPERT F If T the level of print out in TXS file generally is such that only an expert would fully appreciate it QMAINSIM F T if to terminate the simulation at the end of the main period Section 6 1 1 QPT PRIORITY F T if to carry out extension or early recall type of priority signal controls Section 8 4 QTRAJ F True if to output vehicle trajectories in TRJ file Section 9 1 RB GAP 3 0 Acceptable gap sec used upon entering roundabouts Section 6 1 6 RB VISIBILITY 50 Distance on roundabout which the entry traffic would look back for potential conflicts m Section 6 1 6 RDCOLOUR 7 Road colour displayed Section 3 7 2 2 RECALL 5 The amount of recall time of a red signal sec Section 8 4 START HOUR Clock s starting hour to be displa
139. r minute used for calculating generalised costs See Section 9 3 An example of a PAR file is shown below amp PARAM Example CIRC_SPEED 20 NMUC 3 MUC 1 1 Do MUC 2 8 nd MUC 3 16 EI jo amp END Tip Note that the parameters file needs to be created only 1f non default values are to be used For example using the above file the roundabout circulating speed would be reduced to 20 km hr from the default 27 km hr see Section 6 2 There are three user classes in the traffic flow they are of vehicle types CAR TAXI and LGV see Section 6 1 2 for definitions of vehicle type DRACULA Manual v2 4 29 Section 4 Basic Highway Traffic Network and Demand 4 2 Basic Highway Network The network data 1s stored 1n file NET A DRACULA network is represented by zones nodes links and lanes Zones are an abstract concept representing the sources or sinks of traffic where they enter or leave the network They could be used to represent a housing estate a car park or simple connections with traffic from outside the network A node is either external connecting to a zone or an internal intersection connecting to other nodes There is no restriction on the number of roads connected to an intersection A link is a directional roadway between two nodes and consists of one or more lanes A link is specified by its upstream and downstream nodes cruise speed number of lanes and turns permitted to other outboun
140. ream node number of the link where the detector is located Relative position of the lane where the detector 1s located 1 kerbside lane Location of the detector meter If QDEM d maximum recall length seconds Or if QDEM t or r the DETID of the detector to be paired in plan selection signal control Cooling off period seconds if QDEM d ee SECTION 2 ALTERNATIVE SIGNAL DATA mee RECORD 2 1 NODE DATA l 2 3 4 5 6 7 NODE NIN JTYPE QYELL NSTAGE OFFSET Node number Number of links at the node Node type should be 3 for traffic signals a character y or Y entered immediately after JTYPE no space in between JTYPE and QYELL to indicate the junction is in yellow box control Number of stages Relative offset LCY Cycle time for this node seconds kee RECORD 2 2 SIGNAL DATA l 2 3 4 5 STAGL INTG GNA 1 GNC 1 GNA 2 Duration of stage seconds Duration of following inter green seconds The A node for the first turn The C node for the first turn 0 if all turns from A node The A node for the second turn DRACULA Manual v2 4 100 Section 8 Advanced Simulation Applications 6 GNC 2 The C node for the second turn amp DETECTOR E 5 63 1012 50 xampie 6d 2 1345 150 200 E Ei 201T 2 T2 400 202 10 202R 2 T21 200 201 10 99999 amp ALT SIGNAL 7273 3 2 0 20 35 5 731 723 728 15 5 731 723 724 99999
141. restriction The only requirements are a different field being separated by a space unless specified otherwise and b each record type being entered on a separate line with a line break immediately following the last data entry of the record There should be no white space s at the end of a record 4 2 1 1 File Name Specification File names in this section must be preceded by a single line containing the string amp FNAME and terminated by a single line containing 99999 The names should be entered one per line The following files can be specified this way are listed in Table 2 6 marked with an Counter clockwise under drive on the right DRACULA Manual v2 4 31 S ection 4 Basic Hiehway Traffic Network and Demand An example of filename specification is shown below amp FNAME FILTRP ottrip 1 FILBUS otbus 99999 amp FNAME Example FILTRP ottrip1 trp O FILBUS otbus bus 99999 nos E 4 2 1 2 Node and Link Data Formats Data in this section must be preceded by a single line containing the string amp LINKS and terminated by a single line containing 99999 For internal nodes all three record types of Section 2 data are required to give a complete description of the nodes its links and signal timings for traffic signal control intersection For external nodes only those starred variable NAME s of record types 1 and 2 need to be included DATA NAME
142. s been waiting to find an acceptable gap as a stimulus to induce use of a reduced gap The reduced gap is modelled as a linear function reducing from the normal gap GAP to a minimum gap MIN GAP linearly between two waiting times GAP TSTART and GAP TEND Fig 6 2 Acceptable gap sec GAP GAP_MIN Time waited for a normal gap sec GAP_TSTAR GAP_TEND Fig 6 2 Modelling gap acceptance behaviour DRACULA Manual v2 4 84 Example Section 6 Simulation Settings Table 6 4 Global gap acceptance parameters Parameters Defaults Definition GAP Normal acceptable gap seconds 1 0 GAP MIN Minimum acceptable gap seconds GAP TSTART GAP TEND RB GAP Stimulus to induce use of reduced gap time waited for a normal gap seconds Stimulus to induce use of MIN GAP time waited for a normal gap seconds Acceptable gap seconds used upon entering roundabouts Courtesy factor fraction of traffic who are willing to giveway for minor flow cars or cars wanting to change lane FGW_CAR FGW BUS For gap acceptance at roundabouts RB GAP 1s used instead of GAP A Roundabout limit on how far back on the roundabout circulating track an approaching visibility traffic can view RB VISIBILITY is also introduced Fraction of traffic willing to giveway to buses RB VISIBILITY 50 Distance on roundabout which the entry traffic would look back for potential conflicts metre DRACULA also tries to model
143. service minutes eR RECORD 2 BUS ROUTES l NNODES Number of nodes en route See Note 11 2 NODE 1 First node en route 3 NODE 2 Second node en route etc etc 11 NODE 10 The 10 node en route Start with a new line 12 NODE 11 The 11 node en route etc etc ee RECORD 3 LIST OF BUS STOPS l NSTOP Number of bus stops en route See Note 111 2 STOP 1 First bus stop for the service 3 STOP 1 Second bus stop for the service etc etc 4 STOP 10 The 10 bus stop Start with a new line 11 STOP 11 The 11 bus stop etc etc NOTE a If TRROUTE T or t and the node list is A Z then a return bus route with the nodes in exactly reversed order e g Z A is automatically created by DRACPREP and added to the bus route list If however TRROUTE R or r then only one route is assumed whose nodes are Z A The R or r option allows one to code a 2 way route which differs marginally in the two directions Thus if the out direction is A IJL Z and the in direction is Z LKI Z then one may code it using A IKL Z 1 e the same as the out direction apart from one node DRACULA Manual v2 4 T2 T ip Bus route along a straight line Use of SATURN bus routes Section 5 Modelling Public Transport This is a similar function as that exists in SATURN Hence the SATURN users may wish to create the BUS file with the same data as coded in SATURN network data record 66666
144. simulation ends when all vehicles which entered the network within the peak period Section 6 1 1 have completed their origin destination movements The most detailed records at the user s request are the second by second individual vehicles locations and speeds The model also provide point or loop based detector measures on headway distribution flow occupancy and speed For each bus service the model summarizes the mean and standard deviation of total journey time and journey times in between stops a measure which can help distinguish service delay due to traffic congestion from that due to poor management a b Figure 9 1 Simulated traffic conditions at Clifton Green intersection in the City of York Two snapshots were taken at time 08 04 a and 08 20 b as shown by the clock on each snapshot Vehicles are shown as coloured rectangles Part of the network was blocked due to roadworks which are shown in dark grey DRACULA Manual v2 4 111 Section 9 Model Outputs and Evaluation A graphical animation of the vehicles movements can also be shown in parallel with the simulation giving the user a direct view of the traffic condition on the network Two snapshots from DRACULA simulation of an intersection are shown in Figure 9 1 which clearly shows the dynamics in traffic conditions Depending on application DRACULA simulation produces a number of output files These are all ASCII text files the data format
145. sire is triggered The parameters can be set in par file Table 6 5 Parameters specifying lane changing desires and cooling period after one lane changing movement Defaults Fl OVERTAKE 2 0 Threshold speed factor change lane only if the speed can be improved that this factor F2 OVERTAKE Lower boundary factor in desired speed that will trigger a over taking desire TINLANE CAR 15 Time seconds a passenger car keeps in the lane it has just changed into before making another lane changing attempt TINLANE HGV Ditto for heavy goods vehicles Generally drivers make one lane changing movement at a time It takes some time for them to adjust to and to assess the new environment before attempting another lane changing move This behaviour is modelled through a cooling off period assigned to each lane changing movement during that period the vehicle is not permitted to another lane changing move The cooling period is represented with global parameters TINLANE CAR TINLANE BUS and TINLANE HGV which can be specified in par file We model different cooling periods for different type of vehicles to represent the ability of different types of vehicles to move laterally By specifying a cooling period greater than one simulation time step e g 1 second we eliminate chances of any swooping e g a vehicle crosses several lanes at a time such as a sudden change from lane 3 to lane 1 happening in the model DRAC
146. ssigned as link free flow speed should it was coded as zero of blank DRACULA Manual v2 4 41 Section 4 Basic Highway Traffic Network and Demand 4 2 1 7 An example An example coding is given below to describe the network shown in Fig 4 2 with four external nodes two intersections one signalised node 2 and the other priority control node 3 Traffic from link 6 3 gives way to traffic on link 2 3 The route flow for this example is given in Section 4 3 1 Rd ees Pa Ps r P K sages e l 37 3 17 3 sec Fig 4 2 An example network Numbers in rectangles display th node numbers while those in triangles are the zone numbers Junction number 2 is signalised a stage diagram with permitted turning movements and stage length is shown NETWORK dtest net amp LINKS 1 1 0 2 0 2 3 3 2 60 0 1 1 30 0 200 1 1 0 0 3 0 5 1 30 0 200 0 0 1 1 37 3 1 0 E p 17 3 5 0 xam e 3 3 1 e 2 1 30 0 200 1 1 0 0 4 0 6 1 30 0 200 0 0 te q 4 1 0 3 30 0 200 5 i 0 2 0 6 1 0 3 0 99999 amp ZONES i i2 2 4 3 3 5 2 4 6 3 99999 amp COORD 1 300 0 2 100 0 3 100 0 4 300 0 5 00 200 6 100 200 Ci 500 0 C2 500 0 C3 150 250 C4 150 250 99999 DRACULA Manual v2 4 42 Section 4 Basic Highway Traffic Network and Demand 4 2 2 SATURN Network Coding SATURN simulation networks can be used as the base for a DRACULA highway network To start with an existing SATURN ne
147. t tkk RECORD 2 BUS ROUTES l OZONE Origin zone number 2 DZONE Destination zone number 3 NNODES Number of nodes en route 4 NODE 1 First node en route 5 NODE 2 Second node en route etc etc 13 NODE 10 The 10 node en route Start with a new line 14 NODE 11 The 11 node en route etc etc If you have coded the network with this old format you can still use them by setting a parameter QOLDBUS T in your PAR file 5 2 2 Bus stop data Data in this section must be preceded by a single line containing the string amp BUS STOP and terminated by a single line containing 99999 DATA NAME DESCRIPTION l STOPNO Bus stop number ANODE Upstream node number of the link on which the stop lies 3 BNODE Downstream node number of the link 4 POSIT Location of bus stop on link measured from link entry m 5 SIDE Side of the road where the bus stop lies O0 curb side 1 medium side 6 TYPE Type of bus stop 0 ordinary 1 layby J LENGTH Layby length m 8 PFLOW Average pedestrian flow pedestrian hr or average dwell time sec to the bus stop 9 COVPFLOW Coefficient of variation COV of pedestrian flow or COV of dwell time In general public transport means buses including park amp ride buses and guided buses which follow fixed routes in a network If their routes can be specified taxi services can also be coded as a public service in this file But generally taxi trips are included as part of the private trip matrix
148. te the link or network equivalent of speed flow relationships The supply costs reflect the costs experienced by a driver using the network at a given level of demand and they can be used to describe the way in which costs of using net network rise as demand levels increases Since any journey through a network will pass through a number of different traffic states and the costs incurred will be affected by both the journey length and the route taken as well as by the impacts of other demands on the network both at that time and in earlier time periods In order to measure these costs individual vehicles need to be tracked through the network Thus the space time domains used to measure performance curves and supply curves are different as shown in Figure 9 2 and supply curves cannot be readily observed in the way that performance curves can Space A vehicle trajectory for performance curve for supply curve gt Time Figure 9 2 Simulation processing speed versus traffic density It can be seen that if all vehicle trajectories were parallel representing a un congested homogeneous traffic flow performance measures are equivalent to supply measures Performance curves relate the parameters of the traffic in a network at a given time They are based on measurements of vehicle km and vehicle hr in a network in a given time period and can be used to estimate network equivalents of speed fl
149. terval second Average link travel speed kph in the current time period Number of vehicle exiting the link accumulated from the start of simulation e Accumulated average travel time in the current time interval min e Standard deviation of the accumulated travel time min e Accumulated maximum travel time min e Average link travel speed kph If we divide the number of vehicles exiting the link during the current time period by the time period we get a throughput for the link 9 6 Pollution Measures Pollution measures include measures of emission for pollutant CO NOx and HC and measures of fuel consumption Unlike travel condition measures which are measured only when vehicles exit a link or exit the network pollution measures are recorded at every second for each vehicle in the network The program outputs for each link and for the whole network time averages of the pollutant emission and fuel consumption for the current measuring time period say 5 min interval and for the whole simulation time period in the following format and for each link and for the whole network At the end of current time period CO emission in kg for the current output time interval NOx emission in kg for the current output time interval HC emission in kg for the current output time interval Fuel consumption in litres for the current output time interval CO emission in kg accumulated from the start of simulation Accumu
150. tes chosen and records their travel performance At the end of the day the study period a learning model updates the experiences of each individual and stores the information in their travel history files which to a greater or lesser extent influence their next day s choices The current release version code named DRACULA MARS Microscopic Analysis of Road Systems includes only the traffic simulation model and a simplified departure time choice model of DRACULA This version is designed primarily for existing SATURN users who can combine the SATURN route assignment with DRACULA traffic micro simulation for detailed network design and or short term forecasting Hence route assignment is an input exogenous to the model Within DRACULA MARS however there are functions for choice of departure time to be modelled see 4 3 and a wide range of network variability and traffic dynamics These include modelling of a variability in network demand such as daily fluctuation in demand and temporal distribution of demand and supply conditions such as those due to weather and incidents b many dynamic phenomena such as congestion build up blocking back gap acceptance and merging more realistically DRACULA Manual v2 4 l Section 1 Introduction c complex traffic controls such as flared approaches shared lanes and dynamic responsive traffic signal controls d public transport operation and bus priority measures e
151. the Programs 3 5 Running the SPATULA Data Processing SATURN users also benefit from a new program SPATULA which converts DRACSIM time dependent link performance measures to SATURN format See Section 9 2 for details on how to load SPATULA outputs into P1X for data display Step 1 Run SATNET on the network data file to produce UFN file Step 2 Load the program from main menu Module Run SPATULA Step 3 Click Load Network button to load network Or Type the network full path name e g C Dracula net1 spa into the text box Step 4 Start the process by clicking Run button 3 6 Accessing the Programs from SATWIN Front End SATURN users can access the DRACULA MARS programs from within the SATURN Launch Pad SATWIN Select menu SATWIN Module Run DRACULA which will lead you to the DRACWIN front end Follow the instructions in Section 3 2 3 5 to access the DRACULA MARS programs 3 7 Running the Programs in Batch Mode Alternatively the programs can be run with two commands from the main menu Command Line Run Prompt or DOS Prompt as DRACPREP netname DRACSIM netname SPATULA netname To load a bitmap image together with the network type in command DRACSIM netname bitmap DRACULA Manual v2 4 20 Section 3 Executing the Programs 3 8 The DRACSIM Window 3 8 1 The Windows Upon starting the simulation you will get a mam parent window with a title bar across the top reading DRACULA MARS
152. tres Default coordinate unit lmetre If XYUNIT 10 then 1 unit of coordinate 10m Same as used in SATURN so can also be specified 1n the network description file NET Section 4 2 1 XYUNIT GONZO Demand factor number of vehicles of all types are factored by GONZO VCPCU 1 0 Per car unit used to convert flow given in pcu or pcu hr to number of vehicles or vehicles hr Default 1 pcu 1 vehicle of type CAR See Sections 4 3 3 1 CIRC SPEED Roundabout circulating speed km hr See Section 4 2 2 3 CIRC RADIUS Minimum radius of roundabout central island metres See Section 4 2 2 3 QRBLANE If T the number of lanes on a roundabout is limited to at least equal to the maximum lanes on entry arms F if the circulating lanes are determined from SATURN roundabout capacity See Section 4 2 2 3 Number of time periods See Section 4 3 1 Number of user classes See Section 4 3 2 NSTEP NMUC DRACULA Manual v2 4 28 Section 4 Basic Highway Traffic Network and Demand Table 4 2 Parameters defining the basic options for DRACSIM Parameter Defaults Definition NSEED Random number seed NSEED2 4321 Random number seed used to generate random arrival headways TMAIN Simulation time period minutes Over written by LTP in SATURN network coding Section 4 2 2 1 e pa Warm up period minutes TWARM TCOOL 10 Cooling off period minutes TOUTPUT Frequency of output statistics minutes PPM 1 0 Pence pe
153. tributions TECHNICAL Remove SATURN bus routes which have zero service frequency A warning message is given in TXP file An error message will be given in TXP file and the program DRACPREP terminated abnormally if a traffic signal stage is coded with more than two links 4 2 1 2 and 4 2 2 3 Auto load BMP and GIS file names specified via SATURN parameters BMPFIL and GISFIL to DRACULA This function will also work for a cordoned SATURN network as long as the BMPFIL and GISFIL are passed to the cordoned network file Output error message and terminate DRACPREP process if the number of A and C nodes specified for traffic signal stage does not match the number required DRACULA Manual v2 4 138 APPENDIXES e Corrected display of signal aspects for turns e Corrected an error in displaying curved links for driving on the right situation e Allow easy specification of a single speed limit on a link by setting the end position of the speed limit as 1 8 3 e Pass new SATURN parameters IFGIS n n 1 2 7 to DRACULA as options to switch certain GIS features on off Defaults are False See section 4 2 2 1 Also introduce them as DRACULA parameters which can be specified in par files Appendix B and add options within DRACSIM windows menu to allow GIS features switched off on 3 8 2 2 e Correction made on the discrepancy between what documented in Section 8 2 and what programmed the marker for the kerb lane in
154. tted to overtake from link 109 101 Example amp RURAL 101 109 63 109 101 1 99999 Note that the permit for overtaking road with or without double white lines is specified for links Hence to make overtaking from both sides of a road possible one needs to code both links in Table 8 2 Parameters specifying rural overtaking behaviour The default values marked are from the TRARR model OGAP 10 Critical gap in opposing traffic seconds Following to desired speed factor 1 15 for Cars Max speed to desired speed 1 2 for HGVs factor FAMAX 1 5 Max acceleration to normal acceleration factor NLENGTH No of vehicle length SRUB Sight distance metre Outputs of the overtaking movements and the total number of overtakings are summarised in file TXS DRACULA Manual v2 4 109 Section 8 Advanced Simulation Applications 8 6 Pedestrian Simulation and Responsive Signal Controls This section describes a methodology developed to simulate individual pedestrians movements around a signalised intersection and their interactions with vehicular traffic The pedestrian walkway networks are represented by pedestrian nodes and links Pedestrians flows and paths are pre determined and inputs to the model pedestrians move along pedestrian links following fixed routes Different pedestrian characteristics are modelled law obeying pedestrians who only cross the road when the green man is on and two types of
155. twork data file rename the SATURN network data file DAT to NET DRACULA uses the simulation network of SATURN only See Chapter 6 of the SATURN Manual for network coding in particular coding for simulation network record 11111 and node coordinates 55555 If there are buffer links in the network you need first to convert them into simulation links within SATURN Hint use SATURN PIX Network editor Not all sections records of a SATURN network data file are used in DRACULA The ones adopted by DRACULA are OPTION SPECIFICATION RECORD PARAMETER SPECIFICATION RECORD SIMULATION NODES AND LINKS RECORD 1 SIMULATION ZONES RECORD 2 NODE AND OR ZONE COORDINATES RECORD 5 FIXED BUS ROUTES RECORD 6 NN BR WN mm Most of the SATURN records on simulation nodes links zones including buffer zones are used straight forwardly by DRACULA in a similar way as they are used in SATURN For a detailed description of SATURN network coding readers are referred to Chapter 5 and 6 of the SATURN Manual van Vliet amp Hall 2002 Within the sections of data adopted the SATURN variables used and their interpretation into DRACULA model variables are given in the following sub sections except for the bus routes data record 5 above which is described in Section 5 Particular attention is given to the adaptation and interpretation of data records which are different from their SATURN use and or are not used by DRACUL
156. uld become aware of traffic flow on other approaches Table 6 6 Parameters specifying response to traffic control at intersection Defaults TAPPRO JNCT 18 0 Estimated time sec before approaching a junction from when a vehicle begins to react to the junction control Estimated based on the current cruising speed Sections 6 1 and XAPPRO JNCT XAPPRO JNCT Distance metres per 30kph upstream of a stop line where a vehicle starts to react to junction control XVIEW JNCT Distance metre upstream of a stop line where a vehicle starts to look for gaps m the conflicting traffic DRACULA Manual v2 4 89 Section 6 Simulation Settings 6 2 Local Settings 6 2 1 Junction or Link Specific Gap Acceptance Parameters SATURN specifies node based gap values in its network description files see SATURN Manual Chapter 6 for details For example in SATURN networks one can specify parameters GAP and GAPR for priority junction and GAPM for merge These node based gap acceptance values are read in by DRACULA and used as the mean gaps for the junction concerned The records in this section define link specific gap acceptance parameters They give the ability to define gap acceptance not only on a junction by junction basis but also approach by approach Wherever link dependent gaps are given they over write the global parameters as defined in PAR file and the node based gaps as specified in SATURN node records To specify gap acceptance para
157. urn marked G gives way to all major turning movements i e those without any marker on an X but shares with other give way movements van Vliet amp Hall 2002 X For opposed right turn which must cross a major flow from the opposite direction This could be a major flow turning right at a priority junction or at a signalised junction F Fora permanent filter at traffic signals which has 100 green all the time M Merge with one other and only one stream of traffic at a priority junction See Note 1 S A turn with a clear nearside exit Used mainly for motorway entry slip roads to indicate the slip road traffic has a clear exit onto the acceleration lane In case of a two lane slip road the first lane traffic has a clear exit onto the acceleration lane and the second lane traffic will merge with the motorway traffic See Note 11 C Used to indicate a turn that has a clear or protected exit to the far side Unlike an S turn a C turn is to the far side and usually takes the far side lane on exit link Can be used for both priority and signalised junction See Note 111 Z Termed as zipped turns to replace a G turn at a priority junction where under congested conditions the priority rule is changed to one goes another goes the two streams of traffic take in turns to merge into the exit link See Note 4 BLANK no opposing flow DRACULA Manual v2 4 36 Section 4 Basic
158. us lane in action a private vehicle would start to try to get off the bus lane 50 metres before the start of the reservation If there is no setback at the downstream end of the link and that is the only lane that permits the vehicle s next junction turning the private vehicle will take the reserved lane on approaching the junction If there is a setback at the downstream end vehicles of all types may use that lane as long as it permits their next junction turning and by moving into that lane will increase their speeds DRACULA Manual v2 4 70 Lane Choice in presence of a reserved lane Section 5 Modelling Public Transport 5 2 Data Files Public transport data are stored in file BUS The file contains three sections of records each is preceded with a line containing a string beginning with a character amp followed by a record name and ended with a line containing a string 99999 The three sections are l Section 1 Bus service routes Preceded by amp BUS SERVICE See Section 5 2 Within this Section there are three types of records 1 Record type 1 Service description 2 Record type 2 Service route 3 Record type 3 Bus stops en route 2 Section 2 Bus stops Preceded by amp BUS STOP 5 2 3 Section 3 Reserved bus lane Preceded by amp BUS LANE 5 2 Within this Section there are two types of records 1 Record type 1 Link description 2 Record type 2
159. used to build a basic highway traffic network model Most of the data are input via text files which can be created or edited with a text editor e g EDIT BRIEF WORDPAD or NOTEPAD It is advisable not to use any word processor for the editing SATURN users can use PMAKE for network editing 4 1 Control Parameters File The control parameters define the options for input data output data and traffic simulation They are stored in an ASCII text file with the same name as that of the network description and extension PAR All parameters have default values so if a parameter is not specified its default value is used If only the default values are used for this run the file does not need to be created hence it being one of the selective inputs see Section 2 2 The parameter should contain at least a beginning and an end identifier PARAMETERS and END Any parameter to be specified must always be set between the two identifiers Any text or space outside identifiers PARAMETERS and END is ignored by the program and so can be used to insert comments Comments can be put on the same line as the parameters but they must be a entered after the parameter definition and b separated from the parameter definition by at least one space Parameter names are case insensitive so for example parameter gonzo is read in as the same as GONZO The parameter values can be logical T or F stands for TRUE or FALSE respectively
160. y second trajectories for vehicles from certain OD pairs following certain routes or randomly selected from all OD pairs DRACULA Manual v2 4 112 Section 9 Model Outputs and Evaluation 9 2 The SPATULA program SPATULA is a program that converts link based simulation results from DRACULA to a format readable by the SATURN graphical display function PIX At the end of a simulation DRACSIM will produce link based performance measures by time period into a text file lt net gt SPA This file is only produced if a simulation ends naturally e g 1t is ended when all vehicles generated in the main demand period Section 6 1 1 have completed their journey This is because the process involved in producing SPA output is quite time consuming Experience suggests that a number of iteration between parameter setting network coding and simulation animation is often required before final numerical outputs are required for analysis To run SPATULA one first needs to run the network through SATNET to produce the SATURN format UFN file The input and output files for program SPATULA are network UFN Output from SATNET network TRP Input DRACULA format routes and route flows network SPA Input DRACULA simulation results network UFD Outputs in SATURN format The diagram below on the left depicts the processes and input output files used Link Data Set DRACSIM net 7 N one of the below ee Lanes stacking capacity Free
161. y to convert the results from their assignment model to the required DRACULA free standing format and store them in text file TRP The format for a single time period assignment is shown below Results from a multiple user class assignment can be represented in this way see also Section 4 3 1 1 For multiple time periods assignment route choice for each time period should be written to one TRP file Section 4 3 1 2 The data records should be preceded by a line containing a string amp ROUTES and terminated with line containing 99999 Users can enter extra comment lines before the preceding and terminating lines As for network coding Section 4 1 2 the entries of data records are in free format only a space is required to separate two data entries DRACULA Manual v2 4 58 Section 4 Basic Highway Traffic Network and Demand DATA NAME DESCRIPTION l OZONE Origin zone number DZONE Destination zone number 3 MUC The user class following this route See 4 3 1 1 4 FLOW Route flow rate trips hr for the subject time period S66 4 3 1 2 6 QSTART A character symbol to indicate the start of the node list en route 7 NODE 1 First node en route 8 NODE 2 Second note en route etc etc 9 QEOL a symbol at the end of a line to indicate continuation of node list on next line No space between the last node on line and the symbol etc etc 10 QEND a symbol to indicate the end of node list No sp
162. yed START MIN 50 Clock s starting minute to be displayed This correspond to the default warm up period TIMECHRG 1 0 Charge rate for time based pricing pence min Section 8 5 TAPPRO JNCT 18 0 Estimated time sec before approaching a junction from when a vehicle begins to react to the junction control Estimated based on the current cruising speed Sections 6 1 and XAPPRO JNCT TBOARDI 4 0 Boarding time per passenger including purchasing a ticket sec person Section 5 2 TBOARD2 1 0 Boarding time for seasonal ticket holders sec person Section 5 2 DRACULA Manual v2 4 130 APPENDIXES Time takes for a bus to open and close door sec Section 5 2 Cooling off period minutes Section 6 1 1 TINLANE BUS 30 Time seconds a bus keeps in the lane it has just changed into before making another lane changing attempt TINLANE HGV Ditto for large goods vehicles TMAIN Simulation time period minutes Section 6 1 1 Time interval min for outputs Section 9 1 TOUTPUT VZTPM XAPPRO JNCT 150 0 Distance metres per 30kph upstream of a stop line where a vehicle starts to react to junction control Sections 6 1 7 and 6 1 8 and TAPPRO JNCT XVIEW JNCT Distance upstream of a stop line where a vehicle starts to look for gaps in the approaching junction Section 6 1 8 Speed km hr below which a Z type turning movement and major flow traffic will take in turns to merge into exit link Section 4 2 1 5 un

User manual - Institute for Transport Studies

Contents

Download Pdf Manuals

Related Search

Related Contents