多变量系统的PID控制器设计  

Design of a PID Controller for Multivariable Systems

作　　者：吴晓威[1] 张井岗[1] 赵志诚[1] 

机构地区：[1]太原科技大学自动化系,山西太原030024

出　　处：《信息与控制》2008年第3期316-321,共6页Information and Control

基　　金：山西省自然科学基金(2007011049);山西省教育厅科技项目(20051311)

摘　　要：针对多输入多输出系统,提出一种基于微粒群算法的灰色系统预测PID控制算法.该方案将多输入多输出系统转化成若干个多输入单输出子系统,使用多变量一阶灰色模型GM(1,N)预测出每个子系统的输出,实现了多输入多输出系统的对角解耦.然后用微粒群算法优化PID控制器的参数,用优化后的PID控制器对每个子系统进行控制.对两个典型的多变量控制过程进行了仿真,结果表明,这种控制方法超调量较小,鲁棒性强,响应速度快,抗干扰能力强.A grey system prediction PID control scheme based on particle swarm optimization is developed for multiinput-multi-output systems. The proposed scheme divides the MIMO （multi-input multi-output） system into a set of multiinput-single-output subsystems. The output of each subsystem is predicted with a multi-variable first order grey model GM（1,N）, and the MIMO system diagonal decoupling is implemented. Parameters of the PID controller are optimized with the particle swarm algorithm, and each subsystem is controlled with the optimized PID controller. Simulations are made on two typical multivariable control processes, and the results show that the presented method is of small overshoot, high robustness, fast response, and strong anti-jamming capability.

关 键 词：微粒群优化算法 灰色预测 多变量PID控制器 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]