A simple stochastic car-following model for traffic flow
The lack of randomness is a drawback of the available car-following model, because the driving behavior is somewhat unpredictable for observers-it should be described in a probabilistic way. For this purpose, we proposed a stochastic car-following model based on the Bando model. According to the psychology knowledge, the human perception-response time (PRT) distribution should be described by the lognormal distribution. Thus, the PRT or the driver sensitivity constant in me Bando model, which is a constant usually, should be substituted by a random variable obeying lognormal distribution. For avoiding the unreasonable phenomena introduced by the nonphysical acceleration, such as the vehicle collisions and the negative speeds, we add some additional monitor mechanism in the model. By performing the numerical simulation under periodic boundary conditions, we find mat traffic jams appear spontaneously without the artificial small disturbance and they persist during all the simulation time in the enough high density regimes. However, traffic jams can never appear in the low density regime where the model is stable.
Jian-ping Meng Li-yun Dong
College of Mathematics, Physics and Information Engineering, Zhejiang Normal University, Jinhua 3210 Shanghai Institute of Applied Mathematics and Mechanics, Shanghai University, Shanghai 200072, China
国际会议
The 5th International Conference on Nonlinear Mechanics(第五届国际非线性力学会议)
上海
英文
1067-1070
2007-06-11（万方平台首次上网日期，不代表论文的发表时间）