Dong Jiuxiang, Wang Youyi, Yang Guang-Hong
College of Information Science and Engineering and the Key Laboratory of Integrated Automation of Process Industry, Ministry of Education, Northeastern University, Shenyang 110004, China.
IEEE Trans Syst Man Cybern B Cybern. 2010 Dec;40(6):1447-59. doi: 10.1109/TSMCB.2009.2039642. Epub 2010 Feb 17.
This paper considers the output feedback control problem for nonlinear discrete-time systems, which are represented by a type of fuzzy systems with local nonlinear models. By using the estimations of the states and nonlinear functions in local models, sufficient conditions for designing observer-based controllers are given for discrete-time nonlinear systems. First, a separation property, i.e., the controller and the observer can be independently designed, is proved for the class of fuzzy systems. Second, a two-step procedure with cone complementarity linearization algorithms is also developed for solving the H( ∞) dynamic output feedback (DOF) control problem. Moreover, for the case where the nonlinear functions in local submodels are measurable, a convex condition for designing H(∞) controllers is given by a new DOF control scheme. In contrast to the existing methods, the new methods can design output feedback controllers with fewer fuzzy rules as well as less computational burden, which is helpful for controller designs and implementations. Lastly, numerical examples are given to illustrate the effectiveness of the proposed methods.
本文考虑了非线性离散时间系统的输出反馈控制问题，该系统由一类具有局部非线性模型的模糊系统表示。通过利用局部模型中状态和非线性函数的估计，给出了离散时间非线性系统基于观测器的控制器设计的充分条件。首先，针对模糊系统类证明了一种分离特性，即控制器和观测器可以独立设计。其次，还开发了一种采用锥互补线性化算法的两步法来解决H(∞)动态输出反馈(DOF)控制问题。此外，对于局部子模型中的非线性函数可测的情况，通过一种新的DOF控制方案给出了设计H(∞)控制器的凸条件。与现有方法相比，新方法能够以更少的模糊规则和更少的计算负担设计输出反馈控制器，这有助于控制器的设计和实现。最后，给出了数值例子来说明所提方法的有效性。
