具有降阶模型的多模型鲁棒自适应控制  

Multi-Model robust adaptive control with reduced model

作　　者：刘家振[1] 王昕[2] 王振雷[1] 钱锋[1] 

机构地区：[1]华东理工大学化工过程先进控制和优化技术教育部重点实验室,上海200237 [2]上海交通大学电工与电子技术中心,上海200240

出　　处：《计算机与应用化学》2010年第1期73-76,共4页Computers and Applied Chemistry

基　　金：国家杰出青年科学基金(60625302);国家自然科学基金资助项目(60704028);上海市基础研究重点项目(07JC14016);长江学者和创新团队发展计划资助(IRT0712);高等学校学科创新引智计划资助(B08021);上海市重点学科建设项目资助(B504);上海市青年科技启明星计划(07QA14015)

摘　　要：传统鲁棒自适应控制由于考虑了实际系统存在的不确定性,在一定程度上扩大了常规自适应控制的应用范围,但是传统鲁棒自适应控制大多只是从系统全局稳定性的角度出发来设计控制器而忽略系统动态和稳态性能,导致其无法在工况多变的实际被控系统中取得令人满意的效果。针对传统鲁棒自适应控制的不足,本文对由ARMA模型描述并包含未建模动态的系统设计了多模型鲁棒自适应控制器。首先采用正则化技术将系统未建模动态转化为系统有界扰动,并在系统降阶模型的基础上根据系统工况的变化设计了多个固定控制器和2个鲁棒自适应控制器,并根据性能指标函数选择最佳控制器作为当前系统控制器以提高系统性能。仿真实验说明当系统存在未建模动态以及系统工况发生变化时,本文设计的控制器能获得较好的控制效果。The traditional robust adaptive control broadens the application of the routine adaptive control because of considering the uncertainty of practice plant.Traditional robust adaptive control solves the problem that the plant would be unstable under some uncertainties,However,it often ignores the dynamic property and steady-state property,which decides that the traditional robust adaptive control is unfit for practice plant,whose working condition changes from time to time.To solve these problems,this paper designs multi-model robust adaptive controller for the plant which can be described by Autoregressive Moving Average（ARMA） model and includes unmodeled dynamics.First,the normalization factor is used to convert the system unmodeled dynamics to bounded disturbance,and then,many fixed controllers and two robust adaptive controllers are designed based on system reduced model according to the range of the working condition and the best controller computed by the performance index function is chosen as the working controller.Simulation illustrates that the proposed method is preferable when there is unmodeled dynamics and changing working condition.

关 键 词：多模型控制 模型降阶 鲁棒自适应控制 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]