基于过去测量信息的仿射T-S模糊系统的有记忆输出反馈控制与状态估计研究

Takagi-Sugeno fuzzy-affine model-based control and state estimation for nonlinear systems has been a hot research area during the past decades. Compared with T-S fuzzy systems with linear local models, the T-S fuzzy affine systems have much improved function approximation capabilities due to the offset terms of their local models. However, it is noted that most of the existing control and state estimation results for T-S fuzzy systems were only derived based on current output measurements, and thus, the designed memoryless controllers and filters tend to be conservative because of the neglect of useful past output measurements information. On the other hand, the past few years also witnessed a growing research interest in novel memory control and state estimation strategies due to their effectiveness in reducing conservatism and improving system performance. Nevertheless, most of the existing memory control and state estimation results were only developed for linear time-invariant systems and there are very few results on memory control and state estimation for complex nonlinear systems..In this project, by using past output measurements and novel nonmonotonic piecewise fuzzy Lyapunov functions, we will focus on the investigation of some key issues on the analysis and design of memory output feedback controllers and filters for T-S fuzzy affine systems in finite frequency domain. Furthermore, this project will also investigate the non-synchronized memory output feedback control and state estimation for T-S fuzzy affine systems under various constraints. The outcome of this project is to establish theoretical foundation on memory control and state estimation methodologies for T-S fuzzy affine model-based nonlinear systems, and in turn, the proposed memory control and state estimation approaches are expected to equip control engineers with more solid analysis and design tools.

基于仿射T-S模糊模型的非线性系统控制与状态估计是控制领域的研究热点之一。目前针对T-S模糊系统的鲁棒控制及滤波研究大多是基于当前时刻测量信息而设计无记忆控制器与滤波器。另一方面，近年来有记忆控制器与滤波器因具有更低的分析与设计保守性和优异的性能而受到越来越多的关注。现有的有记忆控制与滤波研究结果大多是针对普通的线性定常系统，而对于复杂非线性系统的有记忆控制与滤波，尚少有相关成果报道。本项目针对仿射T-S模糊系统有记忆控制与状态估计中存在的若干难点和亟待解决的关键科学问题，拟提出在噪声频域受限情形下，基于过去测量信息及非单调分段模糊Lyapunov函数的有记忆输出反馈控制及滤波方法。在此基础上，进一步研究各类信息综合受限情形下，复杂仿射T-S模糊系统的非同步有记忆输出反馈控制与滤波，以期提出一套较为完整的仿射T-S模糊系统的有记忆控制与状态估计理论及方法，为工程技术人员提供理论与技术支持。

本项目基于过去测量信息以及非单调分段模糊Lyapunov函数，研究了仿射T-S模糊系统的有记忆输出反馈控制及滤波设计问题。在此基础上，进一步提出在噪声频域受限以及系统各类信息综合受限情形下，复杂仿射T-S模糊系统的同步/非同步有记忆输出反馈控制器与滤波器求解算法。本项目的部分理论成果已在机器人、化工等系统的控制与状态估计中进行尝试性应用。本项目的宗旨是提出一套较为完整的仿射T-S模糊系统的有记忆输出反馈控制与状态估计理论及方法，并为工程技术人员提供较为实用的设计方法。基于此项目，在国际SCI期刊上发表论文47篇(其中包括36篇IEEE Transactions系列汇刊论文)，并获授权国家发明专利6项。