A finite state machine model of lane-changing in microscopic traffic simulation

Lane-changing model is a vital component of microscopic traffic simulation. However, the Gipps' framework separates lane-changing from car-following, and neglects driver's intentions. To overcome these weaknesses, a vehicle moving model based on finite state machine (FSM) is proposed in this paper. There are five states in the FSM, and we focus on the lane-changing, 0 behavior. We analyze the condition of intention generation of mandatory and discretionary lane changing, and derive the critical lead gap and follow gap to ensure collision avoidance. Based on the proposed model, a multi-lane traffic simulator is developed. With this simulator, we reproduce the realistic flow-density and lane changing rate-density relation with periodical boundary condition. The results show that the proposed model can regenerate the realistic characteristics of multi-lane traffic flow.