首页 | 本学科首页   官方微博 | 高级检索  
     

基于GA的交叉路口自适应模糊控制器设计
引用本文:王晓薇,王慧. 基于GA的交叉路口自适应模糊控制器设计[J]. 吉林大学学报(信息科学版), 2004, 22(4): 346-351
作者姓名:王晓薇  王慧
作者单位:浙江大学,系统工程研究所,智能交通系统研究中心,浙江,杭州,310027;浙江大学,系统工程研究所,智能交通系统研究中心,浙江,杭州,310027
摘    要:
针对交通领域中模糊逻辑控制器的模糊子集隶属度函数的离散性特点,提出了改进的GA(GeneticAlgorithm)算法,设计了适用于模糊子集隶属度函数编码的带区间范围限制的十进制编码方案,并实现受限分布一致交叉和变异算子.改进的经典赌轮选择法,通过动态确定最优隶属度函数,模糊控制器的控制性能得到提高,避免了病态个体产生,加快了收敛速度.通过一个实例的Matlab仿真结果表明了该方法的有效性.该方法对路口两方向车辆到达率相差较大情况的处理亦收到明显的效果.

关 键 词:交叉路口模糊控制  遗传算法  自适应  隶属度函数
文章编号:1671-5896(2004)04-0346-06
修稿时间:2004-03-20

GA based design of adaptive fuzzy traffic controllers for urban crossroads
WANG Xiao-wei,WANG Hui. GA based design of adaptive fuzzy traffic controllers for urban crossroads[J]. Journal of Jilin University:Information Sci Ed, 2004, 22(4): 346-351
Authors:WANG Xiao-wei  WANG Hui
Abstract:
Since the MF (Membership Functions) of conventional fuzzy traffic signal controllers are generated usually by human operators, and cannot be modified on-line, it is difficult to guarantee the performance of fuzzy control systems to be optimal when traffic flow changed. An adaptive fuzzy traffic controller based on genetic algorithm is proposed. The near optimal membership functions of fuzzy traffic controller were automatically determined by the enhanced GA (Genetic Algorithm) method. A new decimal coding schema was designed to implement the unified coding of fuzzy MF values. An improved selection method, constrained and stepped crossover and mutation operator are presented to avoid the presence of sick-items and speed up the convergence. The effectiveness of the method was shown through Matlab simulation. In reality, it is commonplace that the two directions of a crossroad are extremely distinct, and the proposed method can effectively deal with such cases.
Keywords:fuzzy traffic control  genetic algorithm  adaptive  membership function
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号