Behavioral Simulation: Towards High-Density Network Traffic Studies

A Champion,MY Zhang,JM Auberlet,S Espie

DOI: https://doi.org/10.1061/40630(255)138

2002-01-01

Abstract:Simulation modeling is an increasingly popular and effective tool for analyzing a wide variety of dynamical problems. Road traffic is an example of such problems. Road traffic constitutes a dynamic problem associated with complex processes. These processes are characterized by the interaction of the elements of the system: road users, infrastructures and operators. Traffic can be considered as a supply and demand problem whose difficulty relies on two opposite postulates. The offer responds to a collective use: the road network is dimensioned to allow a certain flow. The demand is individual: each driver wishes to travel under its conditions. Therefore the characteristics of the traffic imply strong constraints for the modeling. Different types of traffic simulation models exist. According to its granularity, a simulation model can be macroscopic or microscopic. A macroscopic model describes the traffic stream, which is represented in some aggregate manner by scalar values of flow rate, density and speed. A microscopic model considers all the vehicles as individuals and the fundamental interactions take the form of mathematical formulas. All these traffic simulation models describe traffic in statistical formats. In addition to these models, the INRETS (French National Institute for Research in Transportation and Safety) has done some research on road traffic simulation based on the real driver behavior for more than ten years and has developed a behavioral simulation model. INRETS' ARCHISIM— like mathematical model—can be used for various traffic studies. However, it has not been widely used. Behavioral simulation need more computing than mathematical simulation: the system is fully distributed (there is no supervisor), the simulation frequency can be very high (from 5hz to 100 Hz) and the environment is precisely modeled (the space unit is millimeter). Therefore the experiments and studies have been mostly done for low-density road networks. The model shows a high potential for the realization of wide range of traffic studies and only the relatively low number of simulated vehicles seems to be a limitation. But today, as computers are getting powerful enough to allow a high number of vehicles, a partnership between INRETS and SRILOG (a company specialized in Intelligent Transport Systems) has been settled and the two partners have jointly decided to validate the model for complex high-density highway networks. In this paper, we first present behavioral road traffic simulation and the simulation model ARCHISIM. Then, we expose a high-density network study — involving a highway section in the Paris outskirts during the morning rush hours—that SRILOG and INRETS have realized with this model. Finally, the encouraging results of the study are presented and discussed.