遗传算法在模糊控制器参数寻优中的应用研究

Application of genetic algorithm to parameter optimization of fuzzy controller

介绍遗传算法的基本原理及实现步骤,并运用于模糊控制器的参数寻优中,对复杂被控对象实现了最优控制。模糊控制规则采用基于解析表达式的模糊数模型,用遗传算法对模糊控制器中的量化因子和比例因子及偏差加权因子进行编码,确定适应度函数,在复制、交叉、变异的进化过程中获得全局最优点,改善模糊控制性能。在Matlab环境中实现遗传算法各算子的编程,并以二阶系统为例,对常规模糊控制器与经过遗传算法寻优后的模糊控制器进行了仿真比较,其结果表明采用遗传算法寻优能获得较好的控制效果,该方法具有良好的动静态特性。

The basic theory and realization of genetic algorithm are introduced,which is applied to the pa-rameter optimization of fuzzy controller for optimal control of the complicated model.The fuzzy control rule adopts fuzzy data model based on analytic expression.The quantization factor,proportion factor and error weight are coded to determine the fitness function,obtain the global optimal point in the process of reproduction,crossover and mutation,and improve the performance of the fuzzy controller.The operators of different Gas are programmed in Matlab environment .With a second-order system as an example the optimized fuzzy controller and the original fuzzy controller are compared,the result shows that the former one has better control quality and performance.