Interval Observer-Based Active Fault Tolerant Control for Discrete-Time Uncertain Switched LPV Systems
Résumé
This paper presents a novel method for co-designing a TNL interval observer and fault-tolerant control (FTC) for discrete-time switched linear parameter-varying (LPV) systems. These systems are subject to faults, unknown but bounded uncertainties, state disturbances, and measurement noise. By introducing weighting matrices T and N, the design gains additional flexibility in calculating the observer gain matrices, ensuring the cooperative condition of estimation errors. The fault is incorporated into an augmented state vector, allowing the TNL interval observer to jointly estimate the lower and upper bounds of both the system state and faults. The FTC is then designed to compensate for the estimated fault and stabilize the closed-loop system in the presence of faults. Sufficient conditions for the stability of the proposed methodology are formulated as Linear Matrix Inequalities (LMIs), derived using the Input-to-State Stability (ISS) property with multiple Lyapunov functions and the Average Dwell Time (ADT) technique. The approach is applied to vehicle lateral dynamics, demonstrating its effectiveness in estimating lateral speed within a tight interval between the lower and upper bounds, while successfully controlling the yaw rate.