基于模糊逻辑的高速公路事件检测算法仿真研究

Simulation and Research on Freeway Incident Detection Algorithm Based on Fuzzy Logic

选取了一段长为1km的双向6车道高速公路为仿真对象,设定仿真模型的全局参数和相关参数后,得到了具体的仿真环境。将模糊规则划分为发生交通事件的规则和不发生交通事件的规则两大类,采用隶属函数来判断交通事件是否发生,推断得到了交通流是否正常的判别公式。选择检测率、误报率和平均检测时间三个最重要的评价指标对仿真结果进行分析,并与几种常用经典算法的性能指标进行对比,发现笔者所建的基于模糊逻辑的高速公路交通事件检测算法是一种综合性能优良的检测算法。

In this paper a 1 km long expressway with dual 3-lane is selected as the simulation object, and specific simulation environment is obtained after setting up of global parameters and correlation parameters of model. The fuzzy rule is divided into two types which are occurrence traffic incident rule and non-occurrence traffic incident rule, and the judgment formula is deduced that traffic flow is whether or not normal by using membership function to infer that traffic incident occurs or not. Three most important evaluation indexes are selected which are detection ratio, false alarm ratio and average detection time to analyze the simulation results, then the result is compared with performance indexes of several typical algorithms. The results show that the freeway incident detection algorithm based on fuzzy is a good detection algorithm with comprehensive properties.