多变量汽温模糊预测控制  

MIMO predictive control based on T-S fuzzy model

作　　者：牛林[1,2] 刘俊勇[3] 

机构地区：[1]成都大学电子信息工程学院,四川成都610106 [2]昆明理工大学国土资源工程学院,云南昆明650093 [3]四川大学电气工程学院,四川成都610065

出　　处：《电力自动化设备》2009年第9期70-72,共3页Electric Power Automation Equipment

基　　金：成都市科技攻关计划项目(07GGYB198SF);四川省教育厅自然科学基金项目(2006C095)~~

摘　　要：针对火电厂锅炉汽温对象具有大迟延、非线性时变的特性,提出一种基于T-S模糊模型的多输入多输出(MIMO)预测控制策略。将MIMO系统分解为多个多输入单输出(MISO)系统,利用T-S模糊模型描述对象的动态特性,模糊规则将非线性对象划分为多个局部子线性模型,并用加权最小二乘法辨识其参数,然后用预测函数原理设计控制器。为提高预测控制性能,采用多步线性化模型构成多步预报器。仿真结果表明对于MIMO系统的长期预报和控制,多步线性化模型预测控制性能优于单步线性化模型预测控制性能。The steam temperature of power plant has the characteristics of nonlinearity,time variation and long delay time, for which a MIMO（Muhi-Input Multi-Output） predictive control based on T-S fuzzy model is proposed. The MIMO system is decomposed into several MISO（Muhi-Input Single-Output） systems. In each MISO system,the T-S fuzzy model is used to approximate the dynamics of nonlinear object and the fuzzy rule is used to divide the nonlinear system into several local linear models. Their parameters are estimated by linear least squares techniques and their controllers are designed by the predictive function principle. To improve the predictive control performance,the multi- step linearization models are used to form the multi -step fuzzy predictor. Simulative results show that, the multi- step linearization of the T- S fuzzy model performs better than the single-step linearization for the long-term prediction and control of MIMO system.

关 键 词：模糊模型 非线性 预测控制 多输入多输出系统 辨识 蒸汽温度控制系统 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]