Actuator fault tolerance based on continuous‐time fuzzy singularly perturbed model

Abstract

AbstractActuator fault tolerance based on fuzzy singularly perturbed model (FSPM) for nonlinear singularly perturbed systems (NSPSs) is investigated. A continuous‐time FSPM with subtractive faults on control input is presented to describe NSPSs with actuator faults, which is less conservative than slow–fast submodel methods. A hybrid fuzzy controller (HFC) consisting of a fuzzy detector and a fault‐tolerant controller can realize real‐time fault tolerance, such that the target system with actuator faults is stabilized. The theorem to solve controller gains is derived by using Lyapunov, singular perturbation, , and Schur complement theories, which can avoid the problem that the feasible solutions cannot be obtained by solving linear matrix inequalities (LMIs). A nonlinear item was replaced by a linear item, which provides a good idea for solving nonlinear matrix inequalities. The sudden and remittent actuator faults of a motor system are suppressed by using the above approaches.