基于元胞自动机原理的微观交通仿真模型

描述了一种对高速路上的交通流仿真和预测的模型.该模型应用了元胞自动机原理对复杂的交通行为进行建模.这种基于元胞自动机的方法是将模拟的道路量离散为均匀的格子,时间也采用离散量,并采用有限的数字集.同时,在每个时间步长,每个格子通过车辆跟新算法来变换状态,车辆根据自定义的规则确定移动格子的数量.该方法使得在计算机上进行仿真运算更为可行.同时建立了跟车模型、车道变换的超车模型,并根据流程对新建的VP算法绘出时空图.提出了一个设想:将具备自学习的神经网络和仿真系统相结合,再根据安装在高速路上的传感器所获得的统计数据,系统能对几分钟以后的交通状态进行预测.

Microscopic Traffic Simulation Mathematic Model Based on Cellular Automata

This paper presents a model for freeway traffic flow simulation and prediction. The model uses cellular automation theory to model complex traffic behavior. The advantage of the cellular automata approach is that the roadway to be modeled is quantized into simple homogeneous cells, time is quanitzed into discrete steps, and physical quantities take on a finite set of values. Also, the state of the cells is updated at each discrete timestep by using a vehicle update algorithm that combines a few vehicle motion models, governed by a relatively small set of parameters. Then vehicles just move one or several cells at each discrete timestep according to the self-defined rule. This approach makes the computer operation feasible. At last, the paper puts forward a suppose that if the simulation system is equiped with the self-study system of NN (neural network) module according to the statistical data from the transducer fixed on the freeway, it can predict the traffic status ahead of 10 minutes.