基于IGA和Hammerstein模型的非线性系统广义预测控制  

Generalized Predictive Control of Nonlinear System Based on Immune Genetic Algorithm and Hammerstein Models

作　　者：杜志勇[1] 王鲜芳[2,3] 

机构地区：[1]河南机电高等专科学校计算机系,河南新乡453002 [2]江南大学通信与控制工程学院自动化所,江苏无锡214122 [3]河南科技学院信息工程系,河南新乡453003

出　　处：《化工自动化及仪表》2009年第3期15-18,共4页Control and Instruments in Chemical Industry

基　　金：河南省教育厅自然科学研究计划项目(2009A480001)

摘　　要：针对非线性程度较高的系统,设计了一种广义预测控制器。该设计基于Hammerstein模型的动静态可分离特性,首先利用具有全局搜索能力的免疫遗传算法(IGA,Immune Genetic Algorithm)在线辨识模型的一些关键参数,然后运用广义预测控制策略实现对该系统的预测控制。仿真试验结果表明,该设计能够准确预测,而且稳定性好、稳态误差小。A generalized predictive controller was designed for a higher nonlinear process. This method was based on the dynamic and static state separable characteristics of Hammerstein model, some key parameters of model were firstly identified on-line by immune genetic algorithm （IGA） which had an ability of global search. Using generalized predictive control strategy to realize predictive control for the system. The result of some experiments simulation shows that the design can predictive accurately and has a better stability and smaller steady-state error.

关 键 词：免疫遗传算法 HAMMERSTEIN模型 非线性 广义预测控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]