CE4211 Traffic Simulation class presentation 2009 [Compatibility Mode]

CE4211 Traffic Simulation class presentation 2009 [Compatibility Mode]

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Traffic Simulation What is it? John Hourdos Transportation Infrastructure Critical Components of Transportation Infrastructure System Drivers Vehicles Roads d highways R d and hi h Freeway system Rural highway system Arterial d A i l and street systems General environment Traffic control devices ITS infrastructures Need tools to design, evaluate, and operate such design evaluate complex systems. John N. Hourdos 2 Thursday, April 16, 2009 Categories of Traffic Analysis Tools Sketch-planning tools/methodologies. Sketch Sk h- l i l/ h d l i Travel demand models Analytical/deterministic tools (HCM-based) (HCMTraffic signal optimization tools Traffic simulation models Thursday, April 16, 2009 John N. Hourdos 3 Travel Demand Models Higher level planning tools tools. Translate land-use and origin/destination survey landinformation to forecasted link volumes. volumes Assume equilibrium conditions. Necessary for urban planning and upkeep. Examples: p TransCAD EMME2 Voyager Thursday, April 16, 2009 John N. Hourdos 4 Traffic Signal Optimization Tools Accelerate and increase accuracy of empirical methods y p originally done by hand. Assume fixed volumes. Not dependent on geometry, driver, or vehicle characteristics. Necessary f developing proposed control alternatives N for d l i d l l i for later evaluation. Examples: T-Plus Synchro John N. Hourdos 5 Thursday, April 16, 2009 Highway Capacity Manual HCM strengths Well tested through field validation. Suitable for isolated facilities with moderate congestion. g Simple to use. Clearly defined data requirements Congestion expanding over a larger area Dynamic evolution of traffic conditions Issues Queues Shockwaves John N. Hourdos 6 Thursday, April 16, 2009 Traffic Simulation Simulation strengths g Able to evaluate dynamic evolution of congestion problems on transportation systems. Division of analysis period y p Evaluate the system as a whole Can capture the variability of driver/vehicle characteristics. Ca eva ate a te at ve des g s beyo d t e c ass ca o es. Can evaluate alternative designs beyond the classical ones. Allow for detailed assessment of traffic management schemes. Evaluation of dynamic impacts (incidents, rerouting, etc.) Data hungry Can be misleading if not calibrated properly g p p y Assume "100-percent safe driving" "100John N. Hourdos 7 Issues I Thursday, April 16, 2009 Applications of Traffic Simulation Models Evaluation h l i E l i at the planning and policy l l d li level Evaluation at the operational level traffic control intersection/urban street operations freeway corridors y ITS Automated Highway Systems (AHS) Public P bli transportation i Optimization John N. Hourdos 8 Design Thursday, April 16, 2009 Applications of Traffic Simulation Models Real time decision support systems Prediction Strategy design/evaluation route guidance g traffic control Research and development New N concepts and algorithms d l ih used independently elements of larger systems Role supply representation dynamic network loading models (DNL) John N. Hourdos 9 Thursday, April 16, 2009 Simulation Models Definition "... the process of designing a model of a real system and conducting experiments with this model for the purpose either of understanding the behavior of the system or of evaluating various strategies (within the limits imposed by a criterion or set of criteria) for the operation of the system" (Shannon, 1975) y ( , ) Discrete event Approaches synchronous Asynchronous Continuous Contin o s time Hybrid John N. Hourdos 10 Thursday, April 16, 2009 Simulation Models Functionality and level of detail y Network representation Flow representation Traffic dynamics Support of control strategies Travel behavior/demand EventEvent-based TimeTime-based Measures of effectiveness (MOE's) John N. Hourdos 11 Overall structure Output Thursday, April 16, 2009 Level of Detail Based on their flow and traffic dynamics representation traffic simulation models are characterized as: Macroscopic Fluid representation of flow Time and space discretization Individual vehicle representation Continuous space Usually discrete time Individual vehicle representation Traffic dynamics through vehicle interactions and movements Many common elements with microscopic y co o ee e sw c oscop c Detailed representation of vehicle dynamics John N. Hourdos 12 Mesoscopic Microscopic Nanoscopic Thursday, April 16, 2009 Overall S uc u e Ove Structure TimeTime- vs. Event-Based Simulation Event Time-based models advance the clock at fixed imeintervals t t may be different for different processes choice of t important for Efficiency Effi i Accuracy EventEvent-based models maintain an event list. First event in the list is processed next sequencing of events may be difficult less control over efficiency John N. Hourdos 13 Thursday, April 16, 2009 Macroscopic Model C a acte st cs ac oscop c ode Characteristics In effect numerical solutions to continuum flow models ff t n m ri l l ti n t ntin m fl m d l Usually deterministic Common for evaluation of freeway corridor operations Traffic dynamics queuing theory kinetic theory simple input/output simple continuum high order continuum Basic approach numerical solution of kinetic equations discretization of space and time John N. Hourdos 14 Thursday, April 16, 2009 Macroscopic Models: Examples FREFLO FREQ Payne, et. al. May, et. al. Freeway management strategies Freeway corridor simulation and optimization (ramp metering) Freeway operations evaluation, incidents, work zones Freeway corridors, ramp corridors metering strategies and traffic control Freeway networks evaluation of alternative network evacuation strategies John N. Hourdos KRONOS Michalopoulos et. al. Papageorgiou et. al. Deterministic, Deterministic Kinematic theory Deterministic, Input/Output analysis Deterministic, Kinematic theory, continuum model Deterministic, continuum model cell transmission model, model hydrodynamic theory traffic flow models 15 METACOR NETCELL Daganzo, et. al. NETVAC Sheffi et al et. al. Thursday, April 16, 2009 Mesoscopic Traffic Simulation Models Flow Representation p individual vehicles or groups of vehicles with similar characteristics (packets) fluid approximation pp queuing theory integrated networks or corridors g linklink-based lanelane-based aggregate by equivalent capacities Detailed timetime-based John N. Hourdos 16 Traffic Dynamics Network Representation Traffic Control Structure Thursday, April 16, 2009 Mesoscopic Simulation Models: Examples p d p CONTRAM DTASQ TRRL Mahut, Fl i M h t Florian operations evaluation, planning, pseudodynamic assignment ITS, ITS DTA ITS operations evaluation control strategy g gy generation, , DTA ITS operations evaluation, control strategy generation, DTA ITS short-term traffic prediction ITS operations evaluation Planning, ITS, DTA packets of vehicles, queue evolution, PC implementation Time-space queuing model Ti i d l several levels of vehicle aggregation, speed-flow relationships and deterministic queuing, distributed implementation, implementation UNIX platforms individual vehicles, speed-flow relationships, UNIX implementation, p individual vehicles speed-density relationships platoon-based movement of individual vehicles, vehicles PC implementation constant link travel times, deterministic queuing, PC implementation DynaMIT, Supply simulator Ben Akiva, Koutsopoulos DYNASMART (DYNASMART-X) Mahmassani DYNEMO Schwerdtfeger PTV INTEGRATION* Van Aerde METROPOLIS DePalma * Microscopic version has also been developed Thursday, April 16, 2009 John N. Hourdos 17 Microscopic T ffi Si l i M d l Mi i Traffic Simulation Models Based B d on the movement of individual vehicles, vehicle b h f i di id l hi l hi l by vehicle, with varying characteristics and multiple classes. Vehicle positions are updated using car-following logic and lane car following changing rules including stochastic components. Variability in driver behavior and vehicle dynamics is explicitly modeled Interactions between vehicles at intersections modeled by righty g of-way, gap-acceptance rules and traffic control logics (pretimed, actuated, adaptive,....) John N. Hourdos 18 Thursday, April 16, 2009 Vehicle Updating At every simulation cycle, the position and speed cycle of each vehicle is updated according to the following logic: if (it is necessary to change lanes) then f( y g ) Apply Lane-Changing Model endif if (the vehicle has not changed lanes) then Apply Car-Following Model endif Thursday, April 16, 2009 John N. Hourdos 19 CarCar-Following Models L n 1 Ln V n (t ) V n 1 ( t ) x (t ) x n 1 ( t ) xn (t) Common Model: Response(t) = sensitivitystimulus(t-T) T: T reaction time i i Stimulus: v (relative speed), x (relative distance) Sensitivity: function of x, speed, traffic conditions y , p , Response: acceleration, speed Thursday, April 16, 2009 John N. Hourdos 20 CarCar-Following Models L n 1 Ln V n (t ) V n 1 ( t ) x (t ) x n 1 ( t ) xn (t) GM models (Herman, Gazis) Vn (t)m (t Vn1(t T) Vn (t T) an (t) l x(t) an (t ) : acceleration of vehicle n at time t , l, m: parameters Thursday, April 16, 2009 John N. Hourdos 21 Lane Changing Models Mandatory and di M d d discretionary l -changing i lanelane h i mandatory: getting off the current lane in order to continue on the d i d path ( i h desired h (e.g. exiting), or to avoid ii ) id lane closure di discretionary: attempting to achieve desired speed, avoid ti tt ti t hi d i d d id following trucks, avoid merging traffic, etc. Thursday, April 16, 2009 John N. Hourdos 22 Dimensions of Lane Changing Start MLC MLC driving conditions not satisfactory driving conditions satisfactory other lanes Left Lane Gap Accept Gap Reject Right Lane Gap Accept Right Lane Gap Reject Current Lane Left Lane Gap Accept Gap Reject Right Lane Gap Accept Right Lane Gap Reject Current Lane current lane Left Current Lane Lane Thursday, April 16, 2009 Left Current Lane John N. HourdosLane Current Lane Current Lane 23 Gap Acceptance Critical Gap if available gap < critical gap: reject the gap if available gap >= critical gap: accept the gap Critical gap is a function of: g p Relative speed First gap Number of gaps rejected Remaining length (mandatory lane changing) Other opportunities Traffic conditions Lead and lag gap gg p A C lag gap lead gap total gap B The Critical Gap is Function of explanatory variables Gng (t) = exp[Xng (t)g + ng (t)] Prob(a gap is accepted) Prob(lead and lag gaps are accepted) lead lead lag lag = Pr(Gtn > Gcr ,tn and Gtn > Gcr ,tn | n ) Thursday, April 16, 2009 John N. Hourdos 24 Examples of Microscopic Simulation Models AIMSUN 2 ANATOLL ARTEMiS ARTIST CASIMIR CORSIM DRACULA FLEXSYT II FREEVU FRESIM HUTSIM INTEGRATION MELROSE MICROSIM MICSTRAN MITSIMLab NEMIS Universitat Politcnica de Catalunya, ISIS and Centre d'Etudes Techniques de l'Equipement University of New South Wales, School of Civil Engineering Bosch Institut National de Recherche sur les Transports et la Scurit Federal Highway Administration Institute for Transport Studies, University of Leeds p , y Ministry of Transport University of Waterloo, Department of Civil Engineering Federal Highway Administration Helsinki University of Technology Queen's University, Transportation Research Group Mitsubishi Electric Corporation Centre of parallel computing (ZPR) University of Cologne (ZPR), National Research Institute of Police Science Massachusetts Institute of Technology Mizar Automazione, Turin Spain France Australia Germany France USA UK Netherlands Canada USA Finland Canada Japan Germany Japan USA Italy y Thursday, April 16, 2009 John N. Hourdos 25 Examples of Microscopic Simulation Models, cont'd PADSIM PARAMICS PHAROS PLANSIMPLANSIM-T SIGSIM SIMDAC SIMNET SISTM SITRASITRA-B+ TRANSIMS THOREAU VISSIM Nottingham Trent University NTU The Edinburgh Parallel Computing Centre and Quadstone Institute for simulation and training Centre of parallel computing (ZPR), University of Cologne University of Newcastle ONERA Centre d'Etudes et de Recherche de Toulouse Technical University Berlin Transport Research Laboratory, Crowthorne T R hL b C h ONERA Centre d'Etudes et de Recherche de Toulouse Los Alamos National Laboratory The MITRE Corporation PTV System Software and Consulting GMBH UK UK USA Germany UK France Germany UK France USA USA Germany A detailed description of microscopic models can be found in: www.its.leeds.ac.uk/smartest Thursday, April 16, 2009 John N. Hourdos 26 Interesting sources on car-following carmodels and simulation i general d l d i l i in l http://141.30.186.11/~treiber/MicroApplet/index.html http://www.microsimulation.drfox.org.uk/index.html h // i i l i df k/i d h l Traffic analysis toolbox web site http://ops.fhwa.dot.gov/trafficanalysistools/toolbox.htm Thursday, April 16, 2009 John N. Hourdos 27 Nanoscopic Simulation Models General characteristics G l h i i detailed driver behavior detailed vehicle dynamics model detailed vehicle-drive interactions vehicle Model Collision Warning Simulator REAMACS (Rear-end Collision Model) RORSIM (Run-offRoad Simulator) AUTOBAHN MIXIC SmartAHS Thursday, April 16, 2009 Organization TRW Ford Battele Thomas Benz Transport Research Center of Rijkswaterstaat PATH John N. Hourdos Country USA USA USA Germany Netherlands USA 28 Modeling Process Transformati on CONCEPTUAL MODEL COMPUTER MODEL Vali dati on Abstraction Vali dati on Verificati on Implementati on NATURAL S YS TEM COMPUTE SOLUTION Vali dati on (Experi mentation) Thursday, April 16, 2009 John N. Hourdos 29 Problems yet unsolved TwoTwo-way left-turn lanes. leftlanes Impact of driveway access. Impact of on-street parking. I f onki Interactions between bicycles, pedestrians, and vehicles sharing the same roadway. Non-p Non-perfect drivers! Crashes Realistic shockwaves Thursday, April 16, 2009 John N. Hourdos 30 ...
View Full Document

Ask a homework question - tutors are online