Decentralized Fuzzy Observer-Based Fault Estimation for Nonlinear Large-Scale Systems

Abstract

This paper presents a fault estimation technique based on the decentralized fuzzy observer for nonlinear large-scale systems, which are considered to be consisted of fuzzy subsystems and uncertain interconnections. Based on a Takagi–Sugeno fuzzy model for the subsystems of the large-scale system and the decentralized fuzzy observer, the $H_{\infty }$ performance of the fault estimation problem is established by using the estimation error model. By using $H_{\infty }$ performance inequalities to address the fault estimation problem, the decentralized fuzzy observer design techniques are proposed to guarantee the fault estimation conditions. Also, sufficient conditions of observer design are converted into the linear matrix inequality formats. Finally, an example is provided to verify the effectiveness of the proposed decentralized fuzzy observer design techniques for fault estimation.

Keywords

• Decentralized control
• estimation error
• fault detection
• fuzzy systems
• large-scale systems
• linear matrix inequalities