基于遗传算法的分数阶PI^λD^μ控制器参数分级整定

A Grade-Adjusting Method of Parameters of Fractional-Order PI^λD^μ Controller Based on Genetic Algorithm

针对交流电机振动抑制问题,提出基于遗传算法的分数阶PIλDμ控制器参数分级整定策略.通过遗传算法对整数阶PID控制器整定,获得一组控制参数;固定这组控制参数,再利用遗传算法对分数阶PIλDμ控制器进行阶次寻优.仿真结果表明,在PID控制器三系数不变的条件下,通过调整控制器的积分与微分阶次,分数阶PIλDμ控制器与整数阶PID控制器相比,动态性能指标均有所改善;所提出的分级整定策略对分数阶PIλDμ控制器参数确定是有效的.在不增加任何系统结构复杂性的前提下,改善了交流电机的振动抑制特性.

A grade-adjusting method of parameters of fractional-order PI^λD^μntroller is presented based on genetic algorithm for the problem of AC motor vibration suppression.Firstly,parameters of integer-order PID controller are determined by genetic algorithm and fixed,and then an optimizing process is applied to the order of fractional-order PI^λD^μontroller with GA.The simulation results show that compared with integer-order PID controller the dynamic performance of fractional-order PI^λD^μntroller is improved by adjusting the integral and differential order of controller under the same condition of the three-parameter of PID controller,which proves that the grade-adjusting is an efficient method for determining parameter of fractional-order PI^λD^μntroller and can improve the vibration suppression performance of AC motor without increasing the system complexity.