摘要
将遗传操作用于模糊规则和控制器参数编码,实现输入变量的合理组合、模糊规则的获取和控制器参数的优化,设计者仅需给出一个运行遗传算法(GA)的个体适应度函数。同时将模拟退火算法(SA)用于优化控制器参数,这种GASA混合优化策略在模糊控制器设计中取得了良好的效果。实例表明了算法的有效性。
The genetic operations are implemented to determine almost all of the parameters in fuzzy controller, such as appropriate combinations of input variables, number of fuzzy rulers and parameters for membership functions. These parameters are encoded into the chromosomes. Setting the fitness function is the only work for the designer of the controllers. A local improvement mechanism, the simulated annealing (SA), is aided to the GA to optimize the parameters of membership functions. And the virtues of combining GA with SA are discussed. An example proves the effectiveness of the proposed approach.
出处
《控制与决策》
EI
CSCD
北大核心
2003年第6期733-735,739,共4页
Control and Decision
基金
国家自然科学基金资助项目(69974008)
21世纪初中国高等教育人才培养体系研究计划资助项目(C369)。
关键词
遗传算法
模拟退火
模糊控制器
混合优化策略
编码
Genetic algorithm
Simulated annealing
Fuzzy logic controller
Hybrid optimization strategy
Encode
