多变量线性时变系统的特征模型及自适应模糊控制方法  

Characteristic Modelling and Adaptive Fuzzy Control Method of MIMO Higher-Order Linear TimeVarying Systems

作　　者：孙多青[1] 吴宏鑫[1] 

机构地区：[1]中国空间技术研究院北京控制工程研究所

出　　处：《宇航学报》2005年第6期677-681,692,共6页Journal of Astronautics

基　　金：国家自然科学基金重点项目(60034010);973计划课题(2002CB312205);国家自然科学基金项目(90405017;10372015)

摘　　要：对于阶数及参数未知的高阶系统,如何设计一个工程上易于实现的低阶控制器,以实现有效的高性能控制,这是目前航天等领域亟待解决的问题。该文采用将对象特征和控制性能要求相结合的方法,对多输入—多输出高阶线性时变系统和一类非线性时变系统从理论上详细推导出了其特征模型,在此基础上设计了基于特征模型的稳定的自适应模糊广义预测控制方案。由于控制结构中使用了分层模糊系统,从而减少了模糊规则数目。通过对一个挠性航天器控制的仿真实验验证了所给方法的有效性。For higher-order systems with unknown order and parameters, there are still some key problems to be solved in the field of astronautics at present, such as how to design lower order controllers in order to realize effective control with high performance. By incorporating the characteristics of plants with control requirements, 2nd-order characteristic models are deduced theoretically in detail for multi-input and multi-output higher-order linear time-varying systems and a class of nonlinear time-varying systems. On the basis of the results, the adaptive fuzzy generalized predictive control scheme based on characteristic modelling is designed. Because of applying hierarchical fuzzy systems in the control architecture, the number of rules in a （standard） fuzzy controller is reduced greatly. Simulation results on a flexible spacecraft control verified the efficiency of the adaptive fuzzy generalized predictive controller.

关 键 词：挠性航天器 特征建模 线性时变系统 分层模糊系统 广义预测控制 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程] O231[自动化与计算机技术—控制科学与工程]