Robust sensor fault estimation based on a sliding mode observer for linear parameter-varying systems with inexact scheduling parameters

Abstract

This paper presents a sliding mode observer-based sensor fault estimation scheme for multi-affine linear parameter-varying systems with inexact scheduling parameters. A sliding mode related to the fault is maintained by selecting the observer’s discontinuous switching function. Subsequently, the fault is reconstructed using the equivalent output error injection signal. Based on the output error between the system and the observer, an additional nonlinear function is introduced in the observer to address the influence of inexact scheduling parameters on the estimation process. $H_{\infty }$ optimization technology combining with a parameter-dependent Lyapunov function is employed to guarantee the robustness of the observer. By selecting auxiliary variables and using the projection lemma, a sufficient observer synthesis condition is derived as linear matrix inequalities with a single line-search parameter. The numerical examples demonstrate the superiority and effectiveness of the proposed method.