Multi-Agent Based Microscopic Simulation Modeling for Urban Traffic Flow

Abstract

Traffic simulation plays an important role in the evaluation of traffic decisions. The movement of vehicles essentially is the operating process of drivers, in order to reproduce the urban traffic flow from the micro-aspect on computer, this paper establishes an urban traffic flow microscopic simulation system (UTFSim) based on multi-agent. The system is seen as an intelligent virtual environment system (IVES), and the four-layer structure of it is built. The road agent, vehicle agent and signal agent are modeled. The concept of driving trajectory which is divided into LDT (Lane Driving Trajectory) and VDDT (Vehicle Dynamic Driving Trajectory) is introduced. The “Link-Node” road network model is improved. The driving behaviors including free driving, following driving, lane changing, slowing down, vehicle stop, etc. are analyzed. The results of the signal control experiments utilizing the UTFSim developed in the platform of Visual Studio. NET indicates that it plays a good performance and can be used in the evaluation of traffic management and control.

Keywords

• Traffic simulation
• Urban traffic flow
• Multi-agent based system
• Microscopic simulation modelling
• Driving trajectory
• Vehicle driving behaviors.