Fault detection for T-S nonlinear systems with parametric uncertainties via zonotopic H∞ filter

This paper proposes a fault detection scheme via a zonotopic fuzzy $H_{\infty }$ filter for Takagi-Sugeno (T-S) fuzzy systems by considering unknown but bounded parametric uncertainties, disturbances, noises, and actuator faults, which is more consistent with practical systems. First, we design a fuzzy $H_{\infty }$ fault detection filter to obtain robust residuals. The optimal gain matrix of the designed filter is computed offline, which can reduce the computational burden and improve fault detection efficiency. Second, zonotopic analysis is used to obtain the guaranteed adaptive residual thresholds. Third, the residual generation and evaluation scheme through the zonotopic analysis are presented. To illustrate the superiority of the proposed scheme, a numerical simulation comparison is studied. Compared with the existing $H_{-}/L_{\infty }$ method, the proposed scheme has a more concise and simpler design process, which can offer adaptive guaranteed thresholds and detect the fault more accurately. Finally, a vehicle lateral system is used to justify the validity and applicability of the presented fault detection scheme.