基于遗传算法的模糊隶属函数的优化

对于非线性和时变性的不确定系统 ,模糊控制是一种较有效的方法。但在模糊控制器的设计中存在两个瓶颈 :模糊变量隶属函数的确定和模糊控制规则的正确选取。遗传算法是模拟生物的进化现象 ,并采用自然进化机制来表现复杂现象的一种概率搜索方法。传统搜索方法解决不了的复杂问题 ,应用遗传算法却可以得到解决。将遗传算法引入到模糊控制器的设计中 ,根据已知的模糊控制规则 ,对模糊隶属函数自动寻优。以工业过程控制为对象 ,针对具有纯滞后二阶系统进行了计算机仿真 ,并进行了对比性研究。仿真结果表明 ,采用该方法可获得一个基于一定性能指标的最优或次优的隶属函数参数 ,且该模糊控制系统具有过渡过程的时间短、超调小等优点。

Genetic Algorithm Based Optimization of Fuzzy Membership Function

Fuzzy control has been proven to be effective in complex, time-varying and uncertain control systems. But there exist two main problems in the control of complex systems using fuzzy controller: determing and selecting fuzzy control rules and membership functions. Genetic algorithm is a probability-based search method that simulates the evolution process of living beings and adopts a natural evolution machanism to represent complex phenomenon. Unsolvable problem by conventional method can be solved by genetic algorithons. In this paper, genetic algorithm has been used in fuzzy controller design, using this new method, fuzzy membership function can be optimized automatically according to known fuzzy control rules.The results of simulation for second-order system with time-delay show that this method can produce optimal or suboptimal parameters of membership function and give the fuzzy controller some good behaviors.