develop PDF - Read the Docs
Contents
1. r record report S simset Doy y 99 192. e TracNod B node between three modules A TZ PL Figure 2 3 Types of model nodes The vehicle reaching to the node asks for further route even if has saved the whole route Node decides whether or not to pass the vehicle The network node knows only its direct neighbors However its neighbor node can recommend its further neighbor to answer a request The network node can be blocked Putting a lock on the node stops the vehicle despite the fact that the node is only a point on the route Network nodes decide on network traffic All the mechanisms that determine the conduct of network traffic are introduced in the network nodes 2 3 netveh The vehicle moves in a well defined track see laufer Therefore the representation simplifies to the section of the route Due to the fact that the network has a number of vehicles and the fact that they move on the network a vehicle is not a one simple determinated segment VAB direction of motion points describing the object Figure 2 4 Points defining vehicle For safety when the vehicle is not moving we take into account the additional distance before VAD and after him VBC These sections should not be affected by any other vehicle Ifthe vehicle is moving 12 Chapter 2 Code documentation MAULWURF Documentation Release 0 1 11 two addisional distances in the direction of the movement are taken into account VAE VAF Always tracked are two points of
3. 4 We identify all logical nodes on the road and the first from them mark as an event for each of vehicle 5 We create a list of all the events which were identified and take the nearest event 6 We find solution for the taken event depending on the type 7 We update the location of vehicles 1 3 Technical overview and detailed solutions description 7 MAULWURF Documentation Release 0 1 11 8 We move the beginning of the time frame to the just analyzed event time Proper simulation also includes enabling and disabling vehicles moving at the end of their routes In the moments defined by the time step are generated all the changes on the sections of track and they are also saved for future use during time to next time step From a technical point of view the time step is a kind of global event covering both infrastructure and movement by the vehicle Since the operations events on sections of the network are generated in advance for time to next time step the updates that go in the future to the currently analyzed event are ignored in reports associated with the use of vehicles in the network 1 3 5 Global network events New in version 0 1 2 Global events that may occur e time step e introduction of vehicle s to network e dropping network by the vehicle s Global events are plain events with the difference that in general relate to more than one place in the network 1 3 6 The impact of the environment on the network
4. In the simulation it is important to give the bonds of the system with the environment Environment affects the behavior of the system through the network generated events These events are associated with nodes and modules The nodes represent potential points in the network where vehicles can collide with objects outside The collision may occur at longer distance and not be associated with the intersection directions of traffic but their parallelism Then a potential collision is not determined by a network node but the calculated probabilities of an event on module Events on modules of the network coming from the environment are mainly caused by the fact that these sections are not separated from the environment eg damage which is not caused by a network vehicle 8 Chapter 1 User s manual CHAPTER 2 Code documentation Implementation of the model consists of dedicated code modules Python e modules that define the behavior of the model objects e modules that implement the logic and functionality of the model e modules defining the runtime environment and how to cooperate with other software e modules that allow the data logging and reporting for later analysis Currently implemented code modules 2 1 netmod Module NETMOD defines global objects for representation of network s sectors between its nods or allows you to enter and exit the vehicle from the network see sector e TracMod A is a transit sector
5. model DMS Indices and tables mainly only in html genindex modindex Note 1 A link to the current version of MAULWUREF project is always available through http ovzo net To find it please navigate to Projekty Project website 2 The documentation is written in English to make it publicly available it is not my native language If you find any language mistakes or errors don t complain please send me a bug report or make the corrections yourself The document consists of three parts user s manual code documentation python modules attach ments Contents 1 MAULWURF Documentation Release 0 1 11 2 Contents CHAPTER 1 User s manual Description of installation configuration and maintenance of the library and tools 1 1 Installation and configuration of the package 1 1 1 System requirements The requirements of the package are e Python 3 3 e Ipython 1 1 e Sphinx 1 2 1 1 2 Components and architecture Warning The package includes only a few usage examples of the library and it should be so After successfully installation please go to your work directory and run ipython engine to access the library You can find some examples of ipython netbook session in the repository 1 1 3 Sources Sources of this project are maintened in GIT repository public GITHUB private development gito lite You can find the main website of the project on https sec ovzo net d
6. points describing the main events ie EAD EDD EBE i EAC i d ps H i EAE i 5 EAB i e a Figure 2 5 Types of events occurring on the module Event times decide when the network status updates during the simulation Static events are calculated at the beginning of the simulation or during time step initialization Dynamic events are updated on a regular basis during the simulation So the cause of the event can be a vehicle and infrastructure eg as a result of the module EFECTS or current simulation herself eg to allow monitoring 2 4 events 13 MAULWURF Documentation Release 0 1 11 Generally events are logged by network modules but logging through network nodes or vehicles is also allowed 2 5 netini The simulation consists of establishing all the events in a certain period of time The events are mainly related to the routes of vehicles There are two main types of routes e route with closed loop A e a simple route between two points B mmm TRK oes OUT INP Figure 2 6 Route of a vehicle Each of these types consists of three stages e introduction to the network INP e the movement on main route TRK e leave the network OUT Point of entry to the network may be different from the place where vehicle leaves the network Some times you have to take into account the earlier movement between points of entry exit nodes before vehicle is s
7. the vehicle the beginning VA and the end VB This is due to the fact that some event are initiated by the start point of the vehicle and some by its end point These points are therefore recorded on a regular basis If the vehicle is moving additional points VE and VF can occur as source of events The particular event may be related to only one point of the vehicle Various types of vehicles may be present in the network Each one may have different characteristics Vehicles can move at different speeds The value of the average speed on a section is always taken into account Acceleration and braking are not recorded Instead the value of the average speed over some distance is revised therefore it is important to track the VF point This procedure is sufficient even if one vehicle catches up another Distances VAD and VBC are fixed and determined at the beginning of the simulation Distances VAE and VAF are determined and updated on a regular basis They depend inter alia by vehicle type and current speed 2 4 events The basic types of events in the simulation are e events associated with the entry on the module EA and leave it EB e change of the operating parameters on the network section with its lock included static EDD e section reservation for the moving vehicle sometimes including the consequent locking of the whole module dynamic EEE e events at a point static EC stop checkpoint
8. MAULWURF Documentation Release 0 1 11 Piotr Czaja 2014 03 19 Contents MAULWURF Documentation Release 0 1 11 Main goal of this project is to make a set of objects for simulation of guided transit operation systems To achieve this a python library will be made It is intended to be only an analytical tool without graphical user interface I recommend Ipython for use with this library Feel free to use this library for any purpose you like The only restriction is to add link to main page of the MAULWURF project in your paper and software documentation You are welcome to join this project I m ready to consider to grand you write access to the repository after checking some of your earlier proposals and code This document describes both data model for guided transit operation systems DMS and its software reference implementation library SRL Data model consists of network object definitions Each of object descriptions contains e object name the same as classes lowest order DMS SRL e input data designated as Parameters DMS SRL e output data designated as Returns DMS SRL e internal variables designated as Variables SRL The code library is available under license GPL The data model for guided transit operation systems i available under license MIT This document applies to e version 0 1 0 release 0 1 11 of the software package SRL e version 0 1 0 release 0 1 11 of the data
9. el static time frame model Some solutions of network behavior from previous models are still there after adjustment The most simple and natural solution is to divide each segment of the network into smaller equal pieces The procedure resembles in some ways Petri Nets The main problem with this approach is the low accuracy in reflecting lengths The accuracy of measurement is almost the length of the subsegment If two vehicles move along the same path length but a different number of sub segments they can not even meet at a designated place on the network If so you can not take into consideration the exchange of conflict on the network which result from crossing the tracks Treatment pieces of the network as a queue leads to problems if in a single calculation step vehicle traverses many sections Synchronization and registration of many events including the dependency between the moving vehicles at different speeds on sections of the network becomes often impossible The next solution is the use of static time frames The time frame is a period in which events are considered The time frame coincides with a major step in the calculation Each time step contains all network events which must be solved including the sequence of events in a single time frame dependency between events 1 3 Technical overview and detailed solutions description 5 MAULWURF Documentation Release 0 1 11 Calculation of each static time frame consis
10. ementation of simulation model and contains main arguments which determined their choice 4 Chapter 1 User s manual MAULWURF Documentation Release 0 1 11 For the network objects description please see the code documentation 1 3 1 Time and synchronization of events in network Changed in version 0 1 2 Figure 1 1 Time frame vs time step time frame is a period of time which establishes the framework for prediction and determination of parameters of the model objects time step is a set of moments in time for prediction and determination of changes in the network infras tructure Infrastructure facilities are determined at intervals specified by the time frame and effects in time steps The properties of the technical objects results of the events that took place in earlier periods time steps The current time frame does not affect the behavior of the infrastructure objects This simplification is due to excessive correlation between events and properties of the network Events related to vehicles are forecasted for the duration of the time frame We solve only the first event during the time frame Then we update movements of vehicles for the moment We move then the beginning of the time frame to this moment On start of simulation time frame coincide with the time step and this repeats every time step moment Other implementations of network behavior used in the past where Petri Nets like model a queue like mod
11. ent maximum speed on network section respectively maximum average indicated e the speed of directly preceding vehicle and its value is the minimum of the three The use of variable speed for single vehicle is not effective in the analysis of the network It significantly increases the computational effort and does not increase the adequacy of the model to the actual network behavior 1 3 3 Events and network nodes Changed in version 0 1 1 Each event corresponds to the appropriate node on the network physical or logically depending on the type of event Each physical node corresponds to a event associated with it Physical nodes are inserted where there are crossovers e crossings e changes in the operating characteristics of track 6 Chapter 1 User s manual MAULWURF Documentation Release 0 1 11 The events are generated mostly by the physical nodes when the end of the vehicle can potentially pass the node during its movement on the network The logical nodes are placed where there are identified events Events that result in the introduction of logical nodes to the network are e catching the vehicle when one of them is moving and the other is stationary e catching the vehicle when they move at different speed e damaged section of track Physical and logical nodes are on two different layers Physical nodes adjacent to sections modules logical nodes are assigned to these sections Both of physical node
12. etting off on a route To start moving vehicles all network objects must be pre initialized in the correct order It is a task of NETINI module consisting of e enable the introduction of load distribution between nodes of a computing cluster enable the introduction of load distribution between processor cores creation of all network objects perform configuration of network objects based on the class of these objects enable the distribution of objects on remote hosts and ensuring the operation synchronization with the root computing node supervision and execution of all other tasks related to the preparation of the network to simulate 14 Chapter 2 Code documentation MAULWURF Documentation Release 0 1 11 After completing all the tasks on the preparation of a simulation module NETINI gives control over the simulation back to SIMSET module 2 6 simset SIMSET module is superior to the other Allows you to run control and supervise the simulation Provides current simulation data about the status of objects modeled network With this module it is possible to introduce visualization during simulation 2 efects EFECTS module contains the whole simulation logic associated with the operation and maintenance of vehicles and infrastructure This module calculates all the events related to the technical degradation of simulated objects 2 8 loging During the simulation informations that should be noted are generated Im
13. ev To clone the repository use Github for releases or major development tasks or the kaminari ovzo net private repository after achieving read write access to it for minor tasks and conceptual work MAULWURF Documentation Release 0 1 11 1 1 4 How to build the package Clone the repository to local folder Checkout into prefered version Move copy documentations Makefile cp docs etc Makefile docs Makefile Build documentation cd docs make html Build package cd root repo directory python3 setup py sdist You will find the new package in sdist directory 1 1 5 How to install MAULWURF library You can use pip 3 x or do it manually python3 setup py install after uncompressing the package directory Please use the git repository it will be not available on PyPi 1 2 Introduction to the library quick start You can use the library in several different ways The simplest one is 1 Make your virtualenv with all the dependencies of the package 2 Prepare the package 3 Install the package Go to your work directory and run ipython s instance 4 5 Check out the installation rerun ipython s sessions examples 6 Make config files for network you want to analyse 7 Make your s simulations experiments 1 3 Technical overview and detailed solutions description You can simulate an network in many different ways This part of documentation describes solutions taken for impl
14. lossary event see Global network events Events and network nodes Time and synchronization of events in network laufer see netveh Speed Vehicles node see Events and network nodes netnod The impact of the environment on the network sector see netmod time frame see Time and synchronization of events in network time step see Simulation steps 3 2 Releases changes roadmap This document applies to version 0 1 0 release 0 1 11 of the project 3 2 1 v0 5 x e Tools for installation testing Input network definition examples Input simulation description examples Ipython sessions examples 3 2 2 v0 4 X e Ipython session example for user s manual e Improving user s manual e Bugfixes and code optimization 17 MAULWURF Documentation Release 0 1 11 3 2 3 v0 3 X e First full functional and executable code 3 2 4 v0 2 x e Release of all objects and basic functionality description e Source code is still hidden 3 2 5 v0 1 x e The model assumptions are established e The library functionality is full defined e The structure of the project and library is stable e Initialization and configuration of project tools 3 3 Python modules GENERATE modindex 3 4 Index GENERATE genindex 18 Chapter 3 Attachments Python Module Index C conect e efects events loging n netini netmod netnod netveh
15. portant informations are generated by very different network objects and events To unify the form of recorded informations and allow analysis based on logs LOGING module is implemented 2 9 report Data analysis is based on information collected during the simulation REPORT module provides access to relevant information from logs and organizes the data recorded during simulation 2 10 conect Objects can be implemented in a single or multiple processes Large load during simulation can also be the cause of a simulation run on a cluster To ensure communication during simulation between objects regardless of differences in e architecture e runtime environment and to allow parallel processing CONECT module is introduced CONECT module is used whenever there is a need for the exchange of information between objects in the simulation This allows objects do not have to deal with an appropriate addressing of other network objects This module allows communication in each case even the exchange of data between processes on dif ferent hosts can be implemented 2 6 simset 15 MAULWURF Documentation Release 0 1 11 2 11 record The purpose of the module is to mediate in writing and reading information during the simulation For network objects does not matter what is used to read and write the data These can be files database or you can use an ORM 16 Chapter 2 Code documentation CHAPTER 3 Attachments 3 1 G
16. r _ anit name session creates blank sector object all parameters required Parameters e name id of the section object e session id of the session values a current sets value from global vari ables b any correct given previously created session id class netmod InOuMod name session Bases netmod Sector Enter exit sector for introducing to or leaving a vehicle the network __init__ name session creates blank sector object all parameters required Parameters e name id of the section object e session id of the session values a current sets value from global vari ables b any correct given previously created session id 2 2 netnod NETNOD module defines the network nodes see node There are three main cases to insert a node to the network e the intersection with a road to show the influence of the environment e splits the route into sections which can occupy only one vehicle including MAB MAC e the mapping of switches in the network For all other cases insertion of a node must be very cautious since the node affects the network traffic In many cases it is more appropriate to place events on the module with the same consequences for the behavior of a particular piece of the network The model allows the use of only two kinds of physical node 2 2 netnod 11 MAULWURF Documentation Release 0 1 11 e DiffNod A node dividing section into two modules
17. s and logical distinguish their specific types because of their different impact on network behavior For the logical nodes a control point may be both start and end of the vehicle during its movement with or without a corresponding additional section before and after it The general principles are e Events include a time and place e Identification of the site is followed by the node e Identification of time is followed by an event e Events define the points in time at which the network state is updated during the simulation What happens in the network between events is for the logic of a network of minor importance and only through the impact on the events itself In the implementation of the model we use only the physical nodes and events We leave the selection of logical nodes as they result from the events 1 3 4 Simulation steps Changed in version 0 1 2 The simulation is done by examining changes in the network for the moment of selected events Time steps their beginnings generates additional events see time step Proper simulation generally consist of the following steps 1 If the beginning of the time frame falls on moment of a time step we analyze the changes in the sections of the network damaged track for period to the next time step 2 We check maximum range during the current time frame for each vehicle 3 We identify the first physical node overcome on the road by each vehicle and mark it as an event
18. the next module is shorter than the vehicle then the vehicle is moved once a reservation on succeeding module is done where the vehicle fit in 1ts entirety Then intermediate modules and nodes on the route are also blocked The entrance to the module can be blocked by e result of blocking MAC section e as a result of events in the section e setting the lock on net section The name of the module consists of three parts the first node the intermediate module the second node Characteristics is described separately for each direction and the direction is determined by the starting node eg module generally 10 23 11 the module with the selected direction 10 23 Module maulwurf netmod class netmod Sector name session This main class defines all basic behavior of network sectors between junctions or other network nodes name class instance property tuple gt allowed methods get 10 Chapter 2 Code documentation MAULWURF Documentation Release 0 1 11 class netmod TracMod name session Bases netmod Sector Transit network sector with FIFO queue _ anit name session creates blank sector object all parameters required Parameters e name id of the section object e session id of the session values a current sets value from global vari ables b any correct given previously created session id class netmod BufrMod name session Bases netmod Sector Buffer network secto
19. ts of the following steps 1 Defining events associated with the operation of the infrastructure on the sections between nodes which are based on the results of the previous periods 2 Determination of the maximum distance that can traverse each of the vehicles 3 Defining of events and related logical nodes 4 Movement of the vehicles when the range goes beyond the following node and within single sections in the events order In each of these cases the ability to block and delay by slower vehicles have to be considered The need of network section s part reservation for each of the vehicles in conjunction with the depen dency in release of space from preceding vehicles causes the correct timing and order of processing events importance The introduction of short time frames to calculations can help to simplify the sce narios Calculation of the static time frames does not reflect all the changes occurring during their lifetime The omitted changes results from the dependencies between paths of different vehicles 1 3 2 Speed Vehicles New in version 0 1 0 Speed of the vehicles is a major determinant of mapping the dynamic behavior of the network at the designated time Constant average speed of vehicles is sufficient to analyze most of the networks Each vehicle can still move at different average speed and be slowed by another preceding vehicle Vehicle speed can not exceed maximum speed of analyzed vehicle e curr
20. with FIFO queue e BufrMod B is a buffer sector with algorithmically generated leaving order e InOuMod C is a enter exit sector for introducing to or leavieng a vehicle the network A B C Figure 2 1 Types of model modules In addition to these basic types of modules it is possible to to distinguish the network modules with fixed or variable direction movement Events that define the behavior of the network are assigned to each sector module We can distinguish the following basic types of events on modules which generates the data as result of related events e vehicle entry to the network section M A e exit the vehicle from the network section M B MAULWURF Documentation Release 0 1 11 occupation section of the network to enter M A MAC stop on the route M Sx checkpoint on the route M Sx damage of section M S2 MS blocking section by a damaged vehicle M Sa M S1 blocking section by a moving vehicle M S2 M S1 change the operating parameters eg slowing of track section M Sa M S1 Some of these events are being logged by the vehicle and some by the section of the track module points describing the object Figure 2 2 The main elements of the module which generates the data as a result of related events If it is necessary to introduce track sections which only one vehicle can enter at a given time that may be achieved by establishing a permanent MC MB Sometimes
Download Pdf Manuals
Related Search