Three-Stage Zonotopic Kalman Filter-Based H_∞ Augmented State Observer for Time-Delay System with Actuator Fault

A novel state estimation and fault detection (FD) method with a three-stage zonotopic Kalman filter (TS-ZKF)-based H_∞ augmented state observer is proposed for a time-delay system with an actuator fault. First, the time-delay system is augmented to decouple the state delay and actuator fault. Second, an observer is designed for the augmentation system by using the H_∞ performance index, and the interval of state estimation is obtained. Then, the size of the zonotope is minimized to further reduce the state estimation residual and to narrow the interval. The actuator fault estimation and its bounds of the original system are extracted from the augmented state estimate vector. Finally, a numerical simulation and a case on the buck-boost circuit both verify the conservatism and accuracy of the proposed algorithm.