基于再励学习与遗传算法的交通信号自组织控制  

SELF-ORGANIZED CONTROL OF TRAFFIC SIGNALS BASED ON REINFORCEMENT LEARNING AND GENETIC ALGORITHM

作　　者：杨煜普[1] 欧海涛[1] 

机构地区：[1]上海交通大学自动化系,上海200030

出　　处：《自动化学报》2002年第4期564-568,共5页Acta Automatica Sinica

基　　金：国家"86 3"项目 ( 2 0 0 1A A4 2 2 4 2 0 - 0 2 )资助

摘　　要：提出一种基于再励学习和遗传算法的交通信号自组织控制方法 .再励学习针对每一个道路交叉口交通流的优化 ,修正每个信号灯周期的绿信比 .遗传算法则产生局部学习过程的全局优化标准 ,修正信号灯周期的大小 .这种方法将局部优化和全局优化统一起来 ,克服了现有的控制方法需要大量数据传输通讯、准确的交通模型等缺陷 .A combinative algorithm of reinforcement learning and genetic algorithm is proposed and is applied to self-organized control of the traffic signals. The reinforcement learning focuses on the optimization of intersection's traffic flow which modifies the split of traffic signal cycle, while the genetic algorithm intends to introduce a global optimization criterion to each of the local learning processes which modifies the cycle itself of traffic signals. This approach overcomes the drawbacks in existing control method such as huge data transfer and communication, accurate traffic model and so on.

关 键 词：再励学习 遗传算法 交通信号 自组织控制 交通系统 信号灯控制 城市交通 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]