Agent-based Traffic Simulation and Traffic Signal Timing Optimization with GPU

With the advantage of simulating the details of a transportation system, the "microsimulation" of a traffic system has long been a hot topic in the Intelligent Transportation Systems (ITS) research. The Cellular Automata (CA) and the Multi-Agent System (MAS) modeling are two typical methods for the traffic microsimulation. However, the computing burden for the microsimulation and the optimization based on it is usually very heavy. In recent years the Graphics Processing Units (GPUs) have been applied successfully in many areas for parallel computing. Compared with the traditional CPU cluster, GPU has an obvious advantage of low cost of hardware and electricity consumption. In this paper we build an MAS model for a road network of four signalized intersections and we use a Genetic Algorithm (GA) to optimize the traffic signal timing with the objective of maximizing the number of the vehicles leaving the network in a given period of time. Both the simulation and the optimization are accelerated by GPU and a speedup by a factor of 195 is obtained. In the future we will extend the work to large scale road networks.