User`s Manual - National Consortium on Remote Sensing in
Contents
1. is capable of producing considerable detail These are before and after screen shots with a generous tolerance of 30 m If you find you are consistently using tolerance parameters other than the defaults you can modify the default values by editing S2P INI For additions and updates to this document please see the S2P web page at www ncgia ucsb edu ncrst research microsimulation or www ncgia ucsb edu vital VEHICLE INTELLIGENCE amp TRANSPORTATION ANALYSIS LABORATORY UC SANTA BARBARA PAGE 92. values at the start and end of the shapes therefore this is the method employed by S2P Divided roads pose special problems They are variously interpreted as a single 2 way road or multiple one way roads The latter treatment is more flexible because it can handle HOV lanes and other special purpose configurations Some vendors code even divided city streets as individual shapes for each traffic direction and turn U turn opportunities then have to be represented as links connecting the carriageways GDT goes so far as to divide these connecting links into 2 shapes Some of these shapes are no more than a couple of meters long This poses a problem for Paramics because very short links are not easily converted into polygonal road tracks that Paramics models Ramps that merge with freeways are given special treatment in Paramics specifying the length of the acceleration lane and move out distance for vehicles already in the freeway S2P detects such ramp connections and codes them appropriately but to do this it must be given the feature code associated with freeways in the shape file A peculiarity of Paramics is that it allows only right hand acceleration lanes in a right hand drive system such as in North America Ramps that join from the left are therefore coded as regular intersections If a road configuration seems entirely intractable it may be resolvable by editing the shape file in a GIS or by editing the output in Paramics VE
3. The broad purpose of generalization is to reduce the amount of data in the Paramics output files Two models of generalization are available Model 1 Douglas Poiker and Circular Arcs Model 1 is based on the Douglas Poiker algorithm and a simple circle fit It is controlled by two tolerance Tolerance parameters Douglas Poiker tolerance is the maximum distance that a straight line link in the Paramics output file may stray from the original GIS polyline position Circular tolerance is the corresponding maximum for a circular arc link in the output file O A With a low Douglas Poiker tolerance output links follow the polyline segments closely and with a tolerance of 0 Figure 2 Original stippled and generalized they are coincident This results in large files with good heavy polyline in which a subset of the positional fidelity Conversely high tolerance results in original shape points are selected small files with longer links that are positionally inaccurate and visually less appealing Moreover line segments tend to meet at sharp angles and this can affect simulation results because vehicles must slow down at these corners Circular arc generalization operates in conjunction with Douglas Poiker generalization in that the software first attempts reduction of the shape to a single circular arc subject to tolerance failing which it applies the Douglas Poiker method subject to tolerance creating multip
4. link in Paramics but since a link could include several hundred polyline coordinates this can create exceedingly large node and link files with downstream implications for computing performance The translation problem is therefore to convert polylines into combinations of straight lines and circular arcs satisfying the competing alternatives of minimizing the number of links while preserving geometric fidelity This is an interesting theoretical problem and there are several possible approaches to a solution S2P version 3 0 implements three methods of data reduction The first applies the well known Douglas Poiker generalization algorithm the second is a simple 3 point arc fit the third is a biarc algorithm Information on these algorithms is contained in the project report available at the project web site Files There are three program files which must reside together in the same folder Executable program Shape2Paramics exe Dynamic link library ShapeDLL dll Initialization file 2P ini The program requests an input Shape file ESRI s ArcView format which consists of 3 files with extensions SHP SHX and DBF The program outputs 6 Paramics files nodes links centres categories linktypes and vehicles To classify each road in its appropriate category based on the GIS data the DBF and Shape file must contain relevant attribute information see Basic Information below Operational Settings There
5. Caltrans Traffic Operations S2P Version 3 0d User s Manual Contract 74A0067 Richard L Church PI Val Noronha Project Director 2003 June 30 VITAL Vehicle Intelligence amp Transportation Analysis Laboratory NCGIA Department of Geography University of California Santa Barbara CA 93106 4060 USA www ncgia ucsb edu Caltrans DOT ECE GIS ITS NCGIA RS UCB UCSB VITAL VEHICLE INTELLIGENCE amp TRANSPORTATION ANALYSIS LABORATORY UC SANTA BARBARA California Department of Transportation Department of Transportation Electrical and Computer Engineering Geographic Information System Intelligent Transportation Systems National Center for Geographic Information and Analysis UCSB Remote Sensing University of California Berkeley University of California Santa Barbara Vehicle Intelligence and Transportation Analysis Laboratory UCSB Glossary PAGE 2 Introduction Traffic microsimulation is a powerful tool that allows planners to simulate the effects of changing travel demand and congestion new routes legal permissions and traffic control strategies Paramics by Quadstone Ltd is one of the leading microsimulation packages available today This software is extensively used in the California Department of Transportation and other agencies to study traffic control issues on freeways and to test planning scenarios At the University of California Santa Barbara UCSB Paramics has been used to simulate emerge
6. HICLE INTELLIGENCE amp TRANSPORTATION ANALYSIS LABORATORY UC SANTA BARBARA PAGE 7 Tutorial Generalization The purpose of this section is primarily to illustrate the effect of the generalization algorithms and parameters Screen captures below are taken from an early test version of S2P where the model choice dialog box is less obtrusive prompting for only the Douglas Poiker and circular tolerance parameters Sample data files used in this tutorial as well as updated code documentation and INI files can be downloaded from the project web page www ncgia ucsb edu ncrst research microsimulation Launch S2P The initial demonstration will not be with a GDT file therefore an alternate INI must be loaded File Load alternate INI Load itms ini from the installation folder Import a GIS file File Open Select the Los Angeles demonstration file itms_la Using the Options Control Parameters menu item apply Model 1 with a Douglas Poiker generalization tolerance of 200 m and circular arc tolerance of 0 Circular arcs are not generated at zero tolerance Generalization will take several seconds Select the Reposition hand icon function from the toolbar Click on State Route 2 near I 210 in the San Gabriel area see Figure 2 Lean on the key for a second to zoom in Both original red and generalized green shapes are visible under sufficient magnification Use a variety of different Douglas Poiker and circular tolerances to observe the e
7. are three aspects of interaction between the user and the software Basic information on the input Shape files coordinate units and field names View control such as zoom and pan Generalization parameters for the curve reduction algorithms 1 Basic Information Coordinate units Shape files come in a variety of coordinate units and the usual shape file set SHP SHX DBF does not specify the coordinate unit This information must therefore be provided to the software A small selection of VEHICLE INTELLIGENCE amp TRANSPORTATION ANALYSIS LABORATORY UC SANTA BARBARA PAGE 4 S2P USERS MANUAL typical units is offered e g meters feet kilometers miles and S2P converts these to its internal working units meters The software detects latitude longitude files from the coordinate domain and offers an option to convert these to UTM coordinates in meters Field names Some of the attribute information that Paramics requires is contained in the DBF fields of the Shape file the Paramics categories file requires information on width speed limit number of lanes and road type urban rural or highway Since these values could be coded into any field in the DBF file S2P must be told where to find them and what values to use as defaults if the data are not available This information is contained in the initialization INI file S2P INI Field names are listed below with a brief explanation These statements are customized for GDT in
8. ffect of the generalization routines on the ereen and blue generalized shapes Figures 3 4 When selecting these parameters in practice the view should be set to high magnification to check for undesirable results with circular arcs and biarcs Click the Douglas Poiker nodes button the second D_P button in the view toolbar to display the selected nodes explicitly VEHICLE INTELLIGENCE amp TRANSPORTATION ANALYSIS LABORATORY UC SANTA BARBARA Rest Ready Figure 4 I 10 at 110 undesirable eastbound arcs left caused by inappropriate parameters PAGE 8 S2P USERS MANUAL Examine the data reduction statistics Statistics Data Reduction Statistics For information examine the names of the fields in the DBF file Statistics Database Field Names Experiment with Model 2 raising and lowering the tolerance and observing the effect on the output Turn on the Biarc nodes in the toolbar to observe where new nodes have been inserted Save the Paramics files File Save As Paramics Files Standard Paramics files nodes links categories centers linktypes vehicles are placed within the selected folder Exit the program Import the files into Paramics ShapeFile tu Paramics File View Options Statistics Help S Bl alaja stela jn 2 2 J sar fecfe 2 ji j r y Needy X 1972035959 Y 404306 2190 Ready X 1972025959 Y 404906 2190 Figure 5 Model 2 biarc
9. le links It is important to monitor results on the screen because it is possible to obtain undesirable results with some combinations of tolerance Generalization operates on each original link at a time Sometimes when links are broken at points other than intersections e g due to change in attribute generalization may result in sudden changes in VEHICLE INTELLIGENCE amp TRANSPORTATION ANALYSIS LABORATORY UC SANTA BARBARA PAGE 6 S2P USERS MANUAL orientation no attempt has been made to correct this in the current version The biarc method below explicitly enforces tangent continuity at such transition points within shapes Model 2 Biarcs The second model computes a sequence of continuous curves biarcs to approximate the path of the polyline This is superior from an esthetic point of view and because the model enforces continuity of tangents between curves it should be expected that traffic flows more smoothly compared with Model 1 In practice this may not be true because vehicles are made to negotiate curves rather than straight lines The controlling parameter for Model 2 is a lateral biarc tolerance stated in meters Whereas Model 1 in the worst case i e tolerance 0 creates a link for every polyline segment its nodes are always a subset of the original shape points therefore it does not exceed the level of detail in the shape file At low tolerances Model 2 can create extremely short links and winding paths Model 2 i
10. mory Allocation The centres file contains a parameter that tells Paramics how much memory to allocate for its operations This number is typically proportional to the number of nodes in the nodes file S2P prompts for a value to be multiplied by the number of nodes to create this memory allocation request If there are problems importing the files into Paramics it may be appropriate to increase this number 2 View Control S2P provides basic zoom pan and positioning functionality Toolbar buttons control zoom in and out and pan in four directions The 1 1 button restores the original display scale The Reposition button hand centers the display at the geographic location where the mouse is next clicked Zoom and pan can also be controlled with and directional arrows on the numeric keypad ShapeFile to Paramics File View Options Statistics Help 5 Qla stelan ee Bise Bar 2 Figure 1 S2P menu showing map view tool bar The eight buttons to the right toggle layers in the display the first and second button in each pair controlling links and nodes respectively The O buttons If the original GIS file polylines and nodes in red The D_P buttons display generalized links and nodes default green while the Arc buttons display circular arcs and nodes default red and the Biarc buttons display Model 2 results blue Only those buttons relevant to the currently active model are enabled 3 Generalization Models
11. ncy evacuation of residential neighborhoods One of the preparatory steps in microsimulation is the creation of a detailed graphic description of the street network showing each traffic lane signals and other traffic control devices and traffic origin zones This is an extremely laborious process partly because of the large scale at which this description must occur Roads are described graphically using straight line and curve sections ensuring that sections of the road have smooth transitions between them Slight variations in alignment slow vehicles down and affect traffic throughput It is fair to say that data entry is a significant impediment to the widespread use of microsimulation Much of the descriptive information required for microsimulation currently exists in GIS files in the custody of state and local Departments of Transportation private sector data vendors and other authorities Unfortunately there is a significant difference between the geometric models used by Paramics and GIS Once this barrier is overcome using an appropriate tool it should be possible to convert large volumes of GIS data into Paramics compatible files With funding from other California Department of Transportation Traffic Operations Division the University of California Santa Barbara undertook to create such a conversion tool For this initial development emphasis was placed on algorithmic issues while usability issues though addressed were not nece
12. p is assigned an elevation of 1 and upper ramps are assigned values of 2 3 4 etc To distinguish between at grade and non grade intersections S2P examines elevations at the extremities of each incident shape The INI file also contains information on preferred display colors and default parameters used by the software PARAM ELEVATIONMULT 6 In databases such as GDT node elevations are ordinal S2P multiplies ELEVATIONMULT by this ordinal elevation to produce absolute elevation in meters which is entered as the z coordinate of the appropriate node Note that this can create fictitious elevations because non zero elevations are assigned only at non grade intersections PARAM RAMPCATEGORY highway minor This is the parameter used to represent ramps in the output categories file PARAM LANEWIDTH 4 This is the width of each lane in meters S2P uses this value to shift some entrance ramps off the centerline node to provide for an acceleration lane GIS files sometimes contain attribute errors and the output of S2P is only as good as the quality of the input data State and local DOTs often the store road information in linearly referenced files It requires extremely accurate centerline coordinate data to translate literally referenced data into link attributes as in the shape file structure S2P is not equipped to do this VEHICLE INTELLIGENCE amp TRANSPORTATION ANALYSIS LABORATORY UC SANTA BARBARA PAGE 5 S2P USERS MANUAL Me
13. put files If another vendor s files are used some of these statements would have to be modified FIELD SPEEDLIMIT Speed 50 The speed limit variable is represented by the DBF field Speed and the default value for speed limit is 50 km h FIELD FREEWAY FCC A15 A25 Values of A15 and A25 in the FCC feature class code field indicate a freeway Roads joining these are treated as ramps FIELD RAMP FCC A63 A ramp is indicated by a FCC value of A63 S2P needs to know if a road is a highway or a ramp to code merge and exit ramps FIELD ROADTYPE null Urban The road type field is used in the categories file Null indicates that road type is not in the data file Urban is the default FIELD LANES Lane 2 GDT codes the number of lanes as the field Lane If this named field cannot be found in the DBF file or if the value is zero the default of 2 is to be used FIELD LONEWAY ONE_WAY FT TF The oneway field indicates whether the road is one way The value FT indicates traffic flow only in the direction in which the link is digitized while a value of TF indicates flow only in the opposite direction FIELD ELEVATION F_ZLEV T_ZLEV The elevation field is used at non grade intersections In many GIS files intersection nodes exist even at non grade intersections In GDT each shape has an elevation at the from and to end with field names F_ZLEV and T_ZLEV respectively Elevations are ordinal not absolute The lowest ram
14. s also liable to create nodes off the path of the original shape Sometimes this is desirable at other times it is not Limitations The output of S2P is only as good as the quality of the input data There are numerous aspects of inaccuracy in street network data including position connectivity and attributes lanes speed limits These are all carried into the Paramics files To some degree S2P can compensate for quality inconsistencies in coordinate data using the generalization models described above Road centerlines described with a low density of coordinates can be enhanced using Model 2 or circular arcs while roads described in too much detail can be generalized using the Douglas Poiker algorithm However all the generalization methods operate on one shape at a time This means that if roads are fragmented into several shapes the software does not join them into longer sections Similarly if the GIS file does not contain nodes at intersections the software will not insert them The following limitations of the conversion process should be noted Freeway overpasses are sometimes represented as at grade intersections in GIS files Typically an associated DBF table then specifies the legal turn permissions at such an intersection and it is only by consulting these tables that one can distinguish between an overpass and a regular 4 way intersection In the case of GDT this determination is made most easily by examining the elevation
15. ssarily perfected in this first attempt Version 3 0 of Shapefile to Paramics S2P creates files for Paramics version 4 Because Caltrans uses street network data from Geographic Data Technologies GDT the software is packaged with default settings specific to GDT data but it is customizable to work with other vendors products As Paramics development continues there will undoubtedly be a need for further refinement of the conversion tool The web sites www ncgia ucsb edu ncrst research microsimulation and www ncgia ucsb edu vital host up to date information on this project VEHICLE INTELLIGENCE amp TRANSPORTATION ANALYSIS LABORATORY UC SANTA BARBARA PAGE 3 Software Features Basic Principles In a GIS a section of road from one intersection to the next i e a link in typical network parlance is typically stored as a connected series of coordinate points i e a polyline The density of points depends on the sharpness of curves and the scale of representation Links may be split where attributes change e g where a 4 lane road narrows into 2 lanes Paramics takes a different approach to representation each network link being either a single straight segment or a single circular arc It is possible to represent a network in Paramics with the same number of links as in a GIS but this ignores link shape and sacrifices positional fidelity Alternately it is possible to treat each segment of each polyline as a straight line
Download Pdf Manuals
Related Search