基于遗传算法的交通信号多层模糊控制模型研究  

Research on Traffic Signal Multi-layer Fuzzy Control Model Based on Genetic Algorithm

作　　者：李瑞敏[1] 陆化普[1] 史其信[1] 

机构地区：[1]清华大学交通研究所,北京100084

出　　处：《武汉理工大学学报（交通科学与工程版）》2009年第3期407-410,共4页Journal of Wuhan University of Technology（Transportation Science & Engineering）

基　　金：国家"十五"重点科技攻关项目(批准号:2002BA404A20B);北京市科技计划项目(批准号:gyykw05070006)资助

摘　　要：针对城市交通信号控制问题,提出了一种基于交通需求强度的单路口多层模糊控制模型,第一层用来判断路口的交通需求强度;第二层为相位优化层,主要完成相位优化功能;第三层为绿灯时间优化层,用来确定各个相位的有效绿灯时间.针对模糊控制模型中隶属度函数优化的问题提出了利用遗传算法进行模糊控制模型中隶属度函数的优化模型,确定了遗传算法各类参数的计算方法.通过与普通多层模糊控制模型仿真的比较,结果表明本文提出的基于遗传算法的多层模糊控制模型具有较好的控制效果.A multi-layer fuzzy controller is developed for intersection traffic signal control based on traffic demand intensity firstly. The multi-layer fuzzy controller includes three layers. The first layer is used to decide the traffic demand intensity of each approach and phase. The second layer is phase optimization layer to determine the phase sequence of an intersection. The traffic demand intensities and phase sequence are then used in the third layer to determine whether the current signal phase should be extended or terminated. Then the genetic algorithm （GA） is used to optimize the member function of the multi-layer fuzzy controller, the architecture of the model based on GA and the parameters of GA are determined. This paper compares the performance of this controller to that without member function optimization at an isolated intersection and the former has better control effect.

关 键 词：交通工程 城市交通信号控制 多层模糊控制模型 遗传算法 

分 类 号：U491.51[交通运输工程—交通运输规划与管理]