Study on comfortable distance based car-following model with trajectory data

Abstract

Car following models usually assume that individual drivers tend to keep a minimum safe distance from the vehicle in front of themselves, or tend to drive at a desired velocity. Through the analysis of NGSIM data, it is found that the distance between leader car and follower car is often not the minimum safe distance when the driver maintains similar velocity with the leading vehicle. Moreover, due to miss perception of leading vehicle's velocity, the following vehicle's desired velocity is not equal to the leading vehicle's, especially when the leading vehicle's velocity changes greatly. In this paper, the comfortable driving distance is proposed to replace the minimum safe distance in the traditional car-following model, and the calculation method of the desired velocity is given. Logistic function was used to improve the linear function in the spring-damper car-following model, which fit the changing of driving velocity of follower vehicles better. Simulation results are provided to validity the model developed in this paper.