Modeling And Simulating The Macro And Micro Traffic Flow In Expressway Based On Multi-Agent System

With the development of Intelligent Transportation System, the multi-agent system technology is widely used in traffic and transportation. Based on the multi-agent system theory, the vehicles and traffic flows in the urban expressway are modeled and simulated. The main work and obtained results are shown as follows:First, the model architecture of the multi-agent system composed of the micro traffic agents is developed. And the author presents a concept of the dynamic critical car-following distance, which can be used to distinguish the vehicle's driving states. Then a method to compute the distance is put forward. After that, the whole lane-changing action is divided into three steps, the author analyzes the lane-changing process and computes the safety conditions.Then, a method to solve the traffic jam in the merging zone is put forward. Its principle is described as follows:(1) Set up a controller agent around that zone to receive the vehicle agents' real-time information; (2) Based on the information, the controller tests if the arriving agents satisfy the safety conditions; (3) If not, the controller will send some instructions to the certain vehicles. In the end, two simulations are made and the results show that this method is effective.At last, the macro multi-agent system model architecture for the expressway system is proposed, then a visual simulation platform for the 3rd Ring Road in Beijing by using Netlogo is set up. In this way, the process of simulation can be shown to the others directly and vividly. In the platform, the discussed micro traffic models are used to control the vehicle agents'movements. And the expressway is divided into different parts, whose states can be perceived by the platform and be shown in the output window at the same time. If the users change the models embedded in it, then the platform will show the traffic results directly.