An H−/L∞ Reduced‐Order Observer‐Based Actuator Fault Detection Scheme for Nonlinear Systems Using T‐S Fuzzy Model

ABSTRACT

This paper addresses the problem of actuator fault detection (AFD) for Takagi–Sugeno (T‐S) fuzzy systems subject to external disturbances. First, the system output is reconstructed via a variable transformation and a reduced‐order technique using the system input, output and output derivative. Based on the reconstructed output, a novel fuzzy reduced‐order observer is then designed. To enhance the sensitivity of the residual to actuator faults and its robustness against external disturbances, design conditions for the proposed fuzzy reduced‐order observer are derived in the form of linear matrix inequalities (LMIs). Unlike existing AFD approaches that can only achieve performance through finite‐frequency techniques, the proposed method allows the actuator fault signal to directly affect the reconstructed system output. As a result, the residual generated by the proposed method exhibits higher sensitivity to actuator faults, while eliminating the limitation of the finite‐frequency methods that require the fault frequency to lie within a given frequency range. This results in superior performance and wider applicability. Finally, the effectiveness of the proposed method is verified through simulations and experiments.