基于滑模的多变量广义预测解耦控制  

Multivariable Generalized Predictive Decoupling Control Scheme Based on Sliding Model

作　　者：戴永彬[1,2] 杨卫东[1] 王少福[1] 张明[1] 

机构地区：[1]北京科技大学信息工程学院,北京100083 [2]辽宁工业大学软件学院,辽宁锦州121001

出　　处：《化工自动化及仪表》2009年第5期18-20,39,共4页Control and Instruments in Chemical Industry

基　　金：国家发改委技术产业项目基金资助(发改办高技1899)

摘　　要：针对多变量控制系统的耦合问题,将广义预测控制和滑模控制结合起来,提出一种基于滑模的多变量广义预测解耦控制方法。首先把m个输入n个输出的多变量耦合系统分解成m个输入单个输出子系统,再通过对子系统输出预测得到滑模切换函数值,求解开环优化求得控制律,最后通过仿真实验表明,该控制方法对多变量耦合系统的控制是正确有效的。For multivariable control systems with coupling,a multivariable generalized predictive decoupling control scheme based on the sliding model was proposed,which combined the generalized predictive control and the sliding model control.The multivariable process with m inputs and n outputs was transformed into that of n sub-system with m inputs and single output.After predicting the system output,the switching function value was predicted,then the control law can be found out by solving an open-loop optimal control problem.The simulation results for multivariable control systems with coupling are given to demonstrate the correctness and effectiveness of approach proposed.

关 键 词：广义预测 滑模变结构 解耦 多变量控制 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]