一种基于多级遗传寻优的模糊控制器设计

模糊控制器设计的核心问题是模糊控制规则表的确定。模糊控制规则主要来源于专家经验和控制工程的知识,因其并不是在任何时候总能获得,这大大限制了模糊控制器的应用前景。提出了一种基于量化因子,比例因子和解析式规则的模糊PID控制器,并采用多级遗传算法对参数进行分组分阶段交替寻优,仿真结果显示该方法取得了良好的效果,在保证控制效果的情况下,大大简化了模糊控制器设计。

Design of a Fuzzy controller based on multilevel GAs

The important issue of designing fuzzy controller is to confirm the rule-table of fuzzy control. But rules of fuzzy controller are learned mainly from experience of experts and cybernetic knowledge. At the same time, not all of them can be obtained. Therefore this limits greatly the application foreground of fuzzy controller. A fuzzy-PID controller is proposed in this paper based on measure gene, proportional gene and rules parsed. And then GAs is used to optimize it's parameters by grouping. The simulation result gives the good control effect. These make it possible that design for fuzzy-PID controller become very easy.