基于微粒群优化算法的单元机组H_∞控制  

H_∞ control of unit plant based on particle swarm optimization algorithm

作　　者：吴忠强[1,2] 肖雪飞[2] 汤雅超[2] 

机构地区：[1]国家冷轧板带装备及工艺工程技术研究中心,秦皇岛066004 [2]燕山大学工业计算机控制工程河北省重点实验室,秦皇岛066004

出　　处：《仪器仪表学报》2013年第9期2051-2057,共7页Chinese Journal of Scientific Instrument

基　　金：河北省自然科学基金(F2012203088)项目资助

摘　　要：针对单元机组提出了一种基于微粒群优化算法的状态反馈H∞控制方案。首先将非线性系统转化为线性时变系统,再把状态反馈H∞控制问题转化为求Riccati方程的解。H∞控制器的控制效果取决于正定矩阵P的选取。为了保证解的存在性,扩大可行解的范围,提高控制效果,将差分进化算法引进到微粒群算法中得到了改进的微粒群优化(PSO)算法,并采用改进的PSO算法求Riccati方程的解,从而实现对控制器参数的实时优化。理论分析和实验结果均表明该设计方案能够得到使系统渐近稳定的状态反馈控制器。由实验结果可以看出该设计方案提高了控制器参数的优化精度,达到了镇定系统和抑制干扰的双重目的,该系统能够在大范围运行工况下稳定运行。In this paper, a kind of state feedback H~ control scheme based on particle swarm optimization algorithm is presented for the unit plant. Firstly, the nonlinear system is transformed into a linear time varying system, then the state feedback H~ control problem is transformed into the problem of solving the Riccati equation. The control effect of the H∞ controller depends on the selection of positive definite matrix P. In order to ensure the existence of the solution,the scope of the feasible solution is expanded,which improves the control effect. An improved particle swarm optimization （PSO） al- gorithm is used to solve the Riccati equation through introducing the differential evolution algorithm,and the real-time opti- mization of the controller parameters is realized. Theoretical analysis and test results show that using the proposed design scheme,an asymptotically stable state feedback H∞ controller can be obtained. The test results also show that the design scheme improves the optimization accuracy of the controller parameters and reaches the double purposes of stabilizing sys- tem and suppressing interference. The system is able to stably operate over a large range working condition.

关 键 词：单元机组 状态反馈H∞控制 粒子群优化 RICCATI方程 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]