基于多变量广义预测控制的不稳定系统控制结构选择方法

A Control Structure Selection Method Based on Multivariable Generalized Predictive Control for Unstable Processes

对于多输入多输出系统,在控制系统设计时首先要对被控变量和操纵变量进行控制结构选择.Bristol提出的相关增益矩阵(Relative gain array,RGA)法,以及学者们后来提出的各种改进方法,都只适用于稳定系统.本文针对不稳定系统,基于多变量广义预测控制(Generalized predictive control,GPC)的闭环控制律提出了一种控制结构的变量匹配准则.通过对预测时域、控制时域等各个参数的优化选择,使系统闭环稳定;由闭环控制律得到被控变量期望值与操纵变量的相关性矩阵,以此得出控制结构的变量配对方案.通过实例研究表明,对于开环不稳定系统,该方法可以得出正确的变量配对结果.

The control structure selection and variables pair- ing are the first step for multi-input and multi-output （MIMO） systems when designing the decentralized control systems. The relative gain array （RGA） method proposed by Bristol and other improved methods are only suit for stable systems. In this paper, a new variables pairing criterion for control structure selection of unstable systems is put forward, which is based on close-loop control strategy of multi-variable generalized predictive control （GPC）. Through optimizing the controller parameters includ- ing predictive horizon and control horizon to make the closed- loop control system stable, the relativity array between inputs and outputs expectations is obtained from the close-loop control strategy, and it can be used to pair the input and output vari- ables to get a proper control structure. Case studies show that the pairing method is correct for unstable systems.