基于模糊理论的交通溢流识别算法

Traffic spillover identification based on fuzzy theory

交通溢流是交通拥堵的一种极端现象,会导致交通系统的严重紊乱。为实现对交通溢流的控制,必须先对其进行识别。以智能模糊推理为理论基础,提出交通溢流模糊识别算法。该算法推理器以车辆排队比率和路段平均速度为输入语言变量,以道路交通溢流严重程度为输出语言变量,采取Mamdani推理法为蕴含规则,在确定基本论域和离散论域的基础上,建立了模糊规则查询表,实现了交通溢流状态的识别。仿真结果表明,该算法的识别正确率达到98%,证明了模糊算法可以较好地实现交通溢流的识别。

Traffic spillover is a kind of extreme traffic jam phenomenon, which can lead to serious disorder of traffic system. Identification algorithm should be carried out first in order to realize traffic spillover control, so the fuzzy identification algorithm was presented based on fuzzy inference theory. The fuzzy identification inference machine took the car queue rate and road average speed as the input language variables, and the traffic spillover severity index as the output variable. Mamdani was adopted as the implication method. Based on the definition of basic language domain and discrete language domain, the fuzzy query rules table was built based on traffic experts＇ knowledge. Finally, the simulation results indicate the correct rate of identification reaches 98%, which proves the fuzzy inference method is a good tool to recognize the state of traffic spillover.