非线性多变量系统的多模型广义预测解耦控制  

Multi-model Generalized Predictive Decoupling Control for Nonlinear Multivariable Systems

作　　者：薛美盛[1] 樊弟[1] 魏衡华[1] 

机构地区：[1]中国科学技术大学自动化系,合肥230026

出　　处：《化工自动化及仪表》2010年第6期5-8,共4页Control and Instruments in Chemical Industry

基　　金：国家高技术研究发展计划课题(2007AA04Z195);国家水体污染控制与治理科技重大专项课题(2008ZX07010-003)

摘　　要：针对实际工业过程中多变量系统存在着非线性、工况范围广、耦合强的特点,提出基于设定值观测器的非线性多模型广义预测解耦控制算法。该方法由线性广义预测控制器、一种新的设定值观测器和切换机构组成。理论分析和仿真结果表明,该控制策略不但可以保证闭环系统B IBO稳定和渐近收敛,而且能够得到很好的控制效果。According to the characteristics of multi-input multi-output（MIMO）,nonlinear and large operating range in the systems of real industrial processes,a new multi-model generalized predictive decoupling（GPC） control algorithm was presented.It was composed of a linear generalized predictive controller,a new reference observer and a switching mechanism.Theoretical analysis and simulation results show that the closed-loop stability and asymptotical convergence of tracking error can be guaranteed,and the control results are improved.

关 键 词：多变量系统 解耦 广义预测控制 观测器 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]