基于分子动力学的车流运行动态特性及其模型

车辆跟驰行为受前导车和道路环境等的影响,将车辆抽象成相互作用的分子,基于分子动力学构建相互作用势函数,建立基于相互作用势函数的分子跟驰模型.采集试验路段不同点位的交通流样本,从视频中获得所需数据,并对加速度波动特性进行分析.将车辆运行状态分为常态行驶,起动加速和减速停车3种,根据实测交通数据对3种车辆运行状态进行模型参数标定,同时对分子跟驰模型进行稳定性分析验证,结果表明,相对于经典GM模型,分子跟驰模型稳定性更好,对实际交通状态拟合程度更高.

Dynamic Characteristics Model of Traffic Flow Based on Molecular Dynamics

Car- following behavior is influenced by leading vehicle and road environment. Vehicles are abstracted into molecules that interact with each other. The interaction potential function is constructed based on molecular dynamics and molecular car-following model is established based on the interaction potential function. The traffic flow samples of different points in the test section are collected to obtain the required data from the video. And the acceleration fluctuation characteristic is analyzed. The three types of vehicles running is introduced into normal driving state, accelerating state and decelerating state. And the measured data are classified based on this. The classified measured data are used to calibrate the parameters of model in different states. And the stability of the molecular car-following model is analyzed. The result shows that the stability of the molecular car-following model is better and the fitting degree of the actual traffic state is higher.