基于初态修正的ILGA控制器参数整定

Application of Iterative Learning Genetic Algorithms Based Initial State Corrections for Parameters Auto-Optimization of Controller

针对迭代学习遗传算法(ILGA)对PID控制器参数进行整定时要求系统初态等于理想初态的问题,采用在迭代前对实际初态自学习逼近理想初态的方法,放松了迭代学习遗传算法的应用条件,并具有迭代学习次数少,整定PID控制器的最佳控制参数快的特点,简化了控制器的设计过程。给出仿真算例显示了该方法在PID控制器参数自动寻优上的有效性。

Aim at the problem of requiring system initial state equal to ideal initial state while optimizing PID controller parameters by iterative learning genetic algorithms（ILGA）, the way of self-learnlng practical initial state approaching to ideal initial state is adopted, which has relaxed the applied conditions of ILGA. The method bears the merits of less iterative times and searching three optimal parameters of PID controller more quickly, then the design of the PID controller is simplified. As the simulation results shown, the effectiveness of the method is verified in optimizing the PID controller parameters.