基于AAFDE-RBF算法的PID参数整定  

PID Parameter Tuning Based on AAFDE-RBF Algorithm

作　　者：刘悦婷 LIU Yueting(School of Media Engineering,Lanzhou University of Arts and Science,Lanzhou 730000,China)

机构地区：[1]兰州文理学院传媒工程学院,甘肃兰州730000

出　　处：《贵州大学学报（自然科学版）》2023年第4期60-65,共6页Journal of Guizhou University:Natural Sciences

基　　金：甘肃省科技计划项目(21YF5GA072)。

摘　　要：针对工业过程中PID(proportional integral derivative,比例积分微分)参数整定难的问题,提出一种自适应调整因子的差分进化算法(adaptive adjustment factor differential evolution algorithm,AAFDE)的神经网络(radial basis function,RBF)方法整定PID控制器的参数。先在差分进化算法中引入自适应调整变异因子,通过定义个体优劣系数引入自适应调整交叉概率因子;再采用AAFDE算法优化RBF的初始参数,建立RBF模型,接着由RBF在线辨识得到梯度信息;最后根据梯度信息对PID的3个参数在线调整。直流电机系统的仿真实验表明,与RBF-PID和DE-RBF-PID相比,AAFDE-RBF-PID控制器动态性能更好、抗干扰性能更强,控制精度更高。Aiming at PID parameter tuning in industrial process,a radial basis function(RBF)method of adaptive adjustment factor differential evolution algorithm(AAFDE)is proposed to tune and optimize the parameters of PID controller.Firstly,the self-adaptive mutation factor is introduced into the differential evolution algorithm DE and the self-adaptive crossover probability factor is introduced by defining the individual merit coefficient.Secondly,the AAFDE algorithm is used to optimize the initial parameters of RBF,establish the RBF model,and then the gradient information is obtained by the RBF online identification.Finally,the three parameters of the PID are adjusted online based on gradient information.The simulation experiment of DC motor system shows that compared with RBF-PID and DE-RBF-PID,AAFDE-RBF-PID controller has better dynamic performance,stronger anti-interference performance and higher control accuracy.

关 键 词：自适应调整因子的差分进化算法 RBF神经网络 PID参数整定 直流电机系统 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]