Abstract: This paper describes the efforts involved in examining the potential effect of electronic toll collection (ETC) as a possible alternative to automatic coin bridges on the Halifax-Dartmouth Bridge in Nova Scotia, Canada. Using microsimulation modeling techniques, a model was created, verified, validated, and applied to determine the impact of ETC on traffic operations. Additionally, the potential costs and benefits with different toll plaza configurations were examined.
Abstract: In this paper we are concerned with the storage and representation of spatiotemporal traffic data required for Intelligent Transportation System applications. Several data needs, specific for Intelligent Transportation System were identified and studied such as; link data, point data, zone data, and movement data. We interfaced a Geographic Information System package (MAPINFO) with a traffic simulator (DYNASMART) to emulate the traffic data collection process from actual systems. Moreover, a special device was built based on a Global Positioning System and a Geographic Information System database and used as a probe for real-time data collection. Data fusion issues are addressed such as the use of a road matching algorithm to filter inaccurate position data. Issues of efficiently storing large quantities of collected time-dependent traffic data are also discussed; an innovative data storage approach is introduced that fits the time-dependent data to polynomial functions. Finally, the issue of visualizing time dependent data is discussed.
Abstract: Das Verkehrsflussimulationsprogramm PELOPS ermoeglicht ein weites Untersuchungsspektrum in bezug auf das Fahrerassistenzsystem ACC. Das hier untersuchte ACC Konzept basiert auf einer Steuerung von Drosselklappe und Bremse, die durch einen konventionellen Regelalgorithmus verstellt werden. In der Regelstrategie sind die maximale Geschwindigkeit, Beschleunigung und Verzoegerung begrenzt. Ein wesentlicher Aspekt der Regelstrategie ist der Bezug zum Fahrverhalten, was einen Regler erfordert, der in hoeheren Frequenzen Regelabweichungen zulaesst. Bisherige Reglerkonzepte sind noch nicht kolonnenstabil, wodurch Geschwindigkeitsaenderungen vorausfahrender Fahrzeuge verstaerkt werden. In langen Kolonnen kann dieses Verhalten zu Unfaellen fuehren. Eine Moeglichkeit, das Stabilitaetsverhalten zu verbessern, ist eine Vergroesserung des Zielabstands, wenn dieser geschwindigkeitsabhaengig ist. Bei stabilen Kolonnen wirkt sich der Einsatz des ACC positiv auf den Kraftstoffverbrauch aus. Neben der Kolonneninstabilitaet ist fuer die Funktionalitaet des ACC wesentlich, dass der Abstandssensor das relevante Hindernis erfasst. Hierbei sollte die Strahlgeometrie des Abstandssensors an die Streckengeometrie angepasst sein, da die Haeufigkeit falsch erfasster Ziele deutlich zunimmt, wenn die Strahlengeometrie unpassend konzipiert ist. In bezug auf die verkehrlichen Auswirkungen kann beim Einsatz von ACC eine Vergleichmaessigung des Verkehrsablaufes festgestellt werden. Dies wirkt fuer den Fahrer komfortsteigernd, da die Haeufigkeit hoher Beschleunigungen und Verzoegerungen abnimmt. Auch die Sicherheitswirkung ist positiv zu bewerten, was sich in einer deutlichen Vergroesserung der "Time to Collision"-Werte auswirkt. Im Gegensatz hierzu kann keine einheitliche Steigerung der Streckenkapazitaet erzielt werden. Nur bei sehr niedrigen Zielabstaenden, die unter den Werten von durchschnittlichen Fahrern liegen, steigt die Durchschnittsgeschwindigkeit bei gleichzeitiger Steigerung der Verkehrsstaerke. Insgesamt ist ACC ein sehr vielversprechendes Konzept, das den heutigen Verkehrsablauf deutlich verbessern kann. Allerdings bedarf es noch weiterer Entwicklungsarbeit, um die sich bietenden Moeglichkeiten ausschoepfen zu koennen. (A) Title in Englisch: Autonomous cruise control - an analysis with the simulation tool PELOPS.
Abstract: The Chicago Central Area Circulator (CAC) is a light rail transit (LRT) system scheduled to serve downtown Chicago by the year 2000. It will operate in its own travel lane parallel to automobile traffic; however, it will interfere with other surface transportation modes at intersections. The traffic and train signal system controlling the interface will be crucial for the successful performance of all modes. The signal control strategy must balance the needs of LRT, buses, automobiles, and pedestrians. For this reason, three LRT priority control strategies were developed. The approach used to analyze train and automobile traffic performance for each of these strategies is described. The CAC design team simulated LRT operation, automobile traffic flow, and intersection control units (ISCs) as the interface between the two modes for all three control strategies. Two different microscopic modeling tools performed the simulation. TransSim II (registered trademark of James R. Hank dba JRH Transportation Engineering) was selected for the transit and signal controller simulation because it realistically models LRT operation. TransSim II (trademark) can also simulate priority strategies, which include arrival time estimation capability for trains and two-way communication between trains and ISCs. TRAF-NETSIM was selected for the traffic flow simulation because of its ability to reproduce traffic conditions, such as individual vehicles, queuing impacts, and potential spillbacks across adjacent intersections. The interface between the simulation programs is signal phasing and timing. This information calculated by TransSim II (trademark) was read into TRAF-NETSIM. The two simulation processes yielded LRT performance measures of speed, travel time, and delay statistics, and automobile performance measures of delay, queue lengths, and spillbacks. This allowed the design team to choose the most appropriate signal control strategy to provide the best overall system performance.
Abstract: As the engineering and planning communities continue their progress toward managed and integrated transportation systems, transit will play an increasing role. Light rail transit (LRT) has already been selected and implemented by 15 U.S. cities as a rail transit alternative. As new or expanded systems are planned and designed, it is essential that engineers have the means to make the best decisions for LRT placement and operations. The purpose of this research study was to investigate the use of the TRAF-Network Simulator (NETSIM) program and JRH Transportation Engineering's TransSim II (trademark) tools for agencies interested in planning and developing LRT systems. NETSIM is one of the few available traffic analysis programs with the flexibility to model the operations and mobility impacts of transit. Similarly, TransSim II (trademark) can model the impacts of transit and has been developed for this purpose. To evaluate NETSIM and TransSim II (trademark) for simulating traffic in pretimed and actuated arterial networks, outputs from the models were compared with real-world field data from Los Angeles and Long Beach, California and Portland, Oregon. The results indicated that the models could produce moderately accurate estimates of field-stopped delay and percent-stops for individual intersections within studied networks. On a systemwide basis, the models produced reasonably reliable, accurate estimates of network travel times and could reproduce most traffic characteristics observed in the field. The models performed well in simulating the control impacts and behavior of LRT in the modeled systems.
Abstract: This article describes Boston's Central Artery/Tunnel (CA/T) Simulator Project. Researchers at the Massachusetts Institute of Technology (MIT) have developed Intelligent Vehicle Highway Systems (IVHS) modeling and simulation capabilities which are now being extended and employed in the CA/T project. Using the microsimulation model to assess the performance of the project's traffic control and surveillance system, the researchers are now evaluating and refining the CA/T IVHS design.
Abstract: This research investigated the use of the Federal Highway Administration's NETwork SIMulation (NETSIM) program and JRH Transportation Engineering's TransSim II (trademark) as a tool for agencies interested in planning and developing light rail transit (LRT) systems. NETSIM is one of the few available traffic analysis programs with the flexibility to model the operations and mobility impacts of transit. Similarly, TransSim II (trademark) can model the impacts of transit and has been specifically developed for this purpose. To evaluate NETSIM and TransSim II (trademark) for simulating and providing accurate descriptive measures of performance for LRT and traffic in pretimed and actuated arterial networks, researchers compared outputs from the models with real-world field data from Los Angeles and Long Beach, California and Portland, Oregon. The results indicated that the models could produce moderately accurate measures of stopped delay and percent stops for individual intersections within studied networks. On a system-wide basis, the models produced reasonably reliable, accurate estimates of network travel times and were capable of reproducing most traffic characteristics observed in the field. They also performed well in simulating the control impacts and behavior of LRT in the modeled systems.
Abstract: This report identifies and recommends measures of impact that are applicable to the operation of at-grade light rail crossings within traffic signal systems. The key point in identifying measures of impact is to maintain consistency with traffic signal measures of impacts. The recommended measures of impact include average delay and queue length. This report illustrates how the analyst can apply both manual calculation methods and computer models to estimate these measures of impact. Included in the discussion is a screening procedure that is designed to minimize total work effort by identifying impacts and mitigating them with the least intensive analysis method. However, if the analysis results are marginal, then full simulation of the traffic signal system including the light rail line is warranted. The recommended programs for such evaluation are TRANSYT-7F for simpler problems and Traf-NETSIM for complex problems and analysis of system variances.
Abstract: This paper describes the evaluation approach that will be used to assess the performance of the Integrated Project Control System (IPCS) for the Central Artery/Third Harbor Tunnel project (CA/T) in Boston. The primary objectives of the evaluation are to assess the impact of IPCS control strategies on traffic flow and support the analysis of critical operational procedures. The framework for the evaluation consists of identifying a set of traffic scenarios and performance measures that will test the robustness of the IPCS. The traffic scenarios are designed to capture nominal and worst case conditions that might occur on the CA/T and range of variability in driver behavior. The primary evaluation tool is a microsimulation model that has been developed to integrate IPCS control strategies with models of traffic flow (and CO generation) on the CA/T.
Abstract: This paper describes the development of a technique for analyzing complex, frequently changing traffic signal timing developed to accommodate adjacent light rail crossings. Signalized intersections located within a few hundred feet of the at-grade light rail transit (LRT) line and high traffic volumes made coordination between the LRT crossing and adjacent traffic signals a necessity. A study of these crossings was performed using TRAF-NETSIM. This study was the final analysis in determining the at-grade feasibility of an LRT Line through this corridor which was also analyzed with TRANSYT-7F and the Highway Capacity Manual Software.
Abstract: The design and implementation of Advanced Traveller Information Systems (ATIS) providing real-time enroute information to drivers should follow insightful analyses into the dynamics of driver decisions and the resulting traffic flow under information to prevent counter-intuitive and counter-productive results. An important yet often neglected aspect of this program is the distribution of benefits both over the driver population and for different origins and destinations in the network. This paper presents modifications to and an application of DYNASMART (Dynamic Network Assignment Simulation Model for Advanced Road Telematics) for this problem. DYNASMART is a simulation framework for ATIS experiments which incorporates: 1) real-time traffic flow and control simulation, 2) dynamic network path processing, and 3) microscopic consideration of driver response to information. A boundedly-rational behavior model is assumed for driver route-choice under non-prescriptive route information. The information strategies are based on multiple paths rather than a single shortest path. Initial paths of drivers were generated from dynamic equilibrium assignments using CONTRAM program and used as input to DYNASMART. ATIS-equipped drivers change their paths based on a behavioral model (with stochastically assigned parameters) and provided information, while unequipped drivers change routes based on self-observation of traffic conditions. The application presented involves the evaluation of ATIS strategies to alleviate traffic congestion due to spectators leaving a major sports event at Anaheim Stadium. A dynamic traffic demand matrix was estimated from partial link-counts. Interesting insights are derived regarding the higher benefits from ATIS to drivers on congested parts of the network. Robustness of the benefits under various information supply strategies and behavioral scenarios are also discussed.
Abstract: In this thesis, the author describes how reasoning can be integrated with perception, how task knowledge can be used to select perceptual targets, and how this selection dramatically reduces the computational cost of perception. These concepts are illustrated for the domain of driving by a robot driving program called Ulysses. A microscopic traffic simulator called PHAROS is developed to define the street environment for this research.
Abstract: This paper describes the Tuen Mun light rapid transit (LRT) system in Hong Kong, together with the FLEXSYT suite of traffic control programs, and its applications. The Tuen Mun LRT system has been operating in the Hong Kong street network since Autumn 1988. At first, there were several accidents, which generated wide and adverse publicity, but they have almost stopped now that everyone has become used to the existence of the system; the new system is very safe. A priority system is used for traffic control, and signal settings are normally calculated on the basis of fixed-time control with averaged flows, although some adaptations are also made to non-fixed-time LRT calculations. The FLEXSYT suite includes: (1) a controller specification (MANDAT); (2) a junction specification (CONDAT); (3) a network specification (NETDAT). The paper describes the use of FLEXSYT to design two example junctions: (1) a junction in the planned LRT system in Croydon, South London; (2) Junction Mj3, which is one of the most complex junctions on the Tuen Mun LRT system. The authors conclude that: (1) although light rail vehicles (LRVs) and road traffic have very different performance characteristics and vehicle sizes, they can share the same road space if due care is taken; (2) designing signals for LRVs requires a rather different from usual approach; (3) capacities should be calculated using FLEXSYT; (4) the communication between LRT equipment and the traffic system must be specified carefully.
Abstract: Since 1986 three methods have been used to evaluate the traffic effects of at-grade light rail transit (LRT) operations in Dallas' North Central Corridor. The objective was to determine the need for and location of any grade separations. The technical data were subsequently entered into the grade separation decision making process that included other factors such as aesthetics, ability to pay, and community opposition or support. The first of the methods calculates the decrease in cross street capacity resulting from the reduction in the progression band caused by preemptive LRT operations. The second method has four modules that estimate the reduction in cross street capacity, the impact of motor vehicle queuing, motor vehicle stopped delay, and reduction in cross street travel speeds resulting from preemptive at-grade LRT operations. The third method estimates the change in various measures of effectiveness by simulating traffic operations with and without priority at-grade LRT operations. The model used for this third method is the TRANSYT-7F traffic signal optimization and simulation model. In addition to these methods, the cost-benefit analysis used for the North Central Line is discussed, along with the potential application of the NETSIM and Traf-NETSIM models.
Abstract: This paper presents some results of the performance analysis of developed beacon-vehicle communication protocols by accurate analysis, using Markov models and computer simulations. It discusses beacon-vehicle communication requirements and methods for communication initiation. It evaluates and compares different methods for media access control. It presents several stochastic models for evaluating the performance of beacon-vehicle communication protocols and determining their optimal parameters. It estimates the expected time duration for the completion of a communication within a limited zone. It calculates the probability of successful recovery after a message collision, allowing for various factors including traffic intensity and vehicle speeds. In particular, the results of a simulation evaluation are presented for the GANTRY protocol, using the SIMCO2 simulation tools, which is based on realistic multi-lane traffic scenarios and modelling. Some objectives of further research are indicated.
Abstract: A description is given of the MONET3 simulator developed to evaluate RTI communications systems. MONET3 simulates a mobile radio network and is used to evaluate complex protocol stacks and performance behaviour under realistic conditions. The four modules, simulator kernel, channel module, event module and evaluation module are described. MONET3 can be used with several different simulators. The combined simulation of MONET3 with the PROMETHEUS simulator Autobahn Simulator (AS) is described. The movement module is outlined. Aspects of AS are considered in more detail including a) the traffic model, b) driver behaviour (longitudinal and lateral), c) vehicle modelling and d) road elements. Simulation results with the communication protocol Framed ALOHA are provided as an example.
Abstract: This paper is intended to stimulate debate about the progress made by inter-urban traffic modelling during the last decade and about its possible future directions. In the UK, the Department of Transport (DTp) has improved the accuracy of the existing models, rather than the models themselves. For urban applications, major improvements have been made by alternative approaches such as SATURN, CONTRAM AND TRAFFICQ, which take explicit account of junction delays, and new methods have been developed for estimating origin-destination flows from traffic counts. For inter-urban applications, equivalent alternative modelling approaches have not yet become established. The paper comments on the current state of the art of conventional forms of inter-urban travel demand, and notes some deficiencies. It indicates some possible ways of improving the present methods, either within the existing modelling framework or with newer techniques. It concentrates on models for estimating origin-destination flows and link flows for private road vehicle traffic, and does not discuss models of vehicle ownership. The author advocates the same healthy scepticism when exploring the use of new techniques as when using previously well tested techniques now known to be inadequate. He suggests that one key to a more useful future for traffic models lies in the user-friendly management of data and models rather than in the models themselves. To achieve this, the concepts of modelling need to be integrated with those of the engineer/computer interface such as computer-aided design and manufacturing (CAD/CAM).
Abstract: Ce rapport presente le programme de simulation microscopique de circulation en file SIMDAC. La modelisation de la conduite en file et le fonctionnement du programme sont exposes dans le premier chapitre. Le second chapitre decrit la mise en oeuvre du modele et le chapitre trois est consacre al' exploitation du logiciel en tant qu' outil de test de dispositifs de signalisation arriere renforcee.
Abstract: The concept of light rail vehicles (LRVs) operating at grade and alternately sharing the right-of-way perpendicular to the flow of automobile traffic is an attractive transit idea because of the potential cost savings to transit agencies. This paper is a partial review of an evaluation of the potential delay impacts on automobile traffic imposed by LRVs operating at grade. This report can assist decision-makers in determining where grade separations are appropriate. Also presented is a methodology for summarizing the operational characteristics of a light rail transit grade crossing with a single parameter, the crossing-volume-to-capacity ratio. The analysis centered on computer simulations using FHWA's NETSIM model. Results indicated that for light rail transit crossings located in excess of 400 ft from any adjacent intersection, the delay imposed on the motoring public warranted a grade separation only at very high traffic volumes or very short LRV headways.
Abstract: The Florida Daprtment of Transportation's Bridge Preemption Study is reported, which attempted to identify the critical factors relating to bridge openings, determine the impact of bridge opening on adjacent signalized intersections, and develop standardized optimization treatment to minimize the impacts. An attempt was made to develop a menu-driven microcomputer program to estimate various measures of effectiveness related to the bridge/signalized intersection network operation. To do this, the NETSIM program was used. A menu-driven, microcomputer-based preprocessor program to generate NETSIM data files for the specific case of a drawbridge located between two signalized intersections was developed. Details are given of the bridge/intersection model. The preprocessor program and NETSIM analyses are described.
Abstract: This paper describes an Automatic Vehicle Monitoring Simulation System (AVMSS). AVMSS is an interactive bus route simulation model designed to act as a test bed for the evaluation of AVM strategies and tactics. The system is comprised of three component programs: a Traffic Environment Generator (TEG), a Bus Schedule Generator ( BSG), and an AVM Simulator (AVMS). The simulation system uses both macroscopic and microscopic simulation techniques. The macroscopic traffic flow model employed in the TEG is based upon the TRANSYT model. Buses are moved in the AVMS using a microscopic time-scanning technique. The features of each component program are discussed. Data requirements and the measures of effectiveness produced are presented. AVM Strategies, imbedded in the model are described.