Data-Based Adaptive Fault Estimation and Fault-Tolerant Control for MIMO Model-Free Systems Using Generalized Fuzzy Hyperbolic Model

Abstract

This paper is focused on the data-driven model-free adaptive fault detection and estimation (FDE) and fault-tolerant control (FTC) problems for multi-input multi-output (MIMO) discrete-time systems with unknown sensor faults. First, in the light of the compact form dynamic linearization method, the initial systems are transformed into a novel data-based model with only one unknown parameter. Second, a fault estimator is established to detect the sensor faults. Noting that a time-varying residual threshold is developed to determine whether the sensor faults occur or not. Then, the unknown faults are approximated based on the powerful approximation capability of a generalized fuzzy hyperbolic model and the FTC approaches are reconstructed by applying the optimality criterion. In contrast to the previous schemes, the main contributions are twofold: first, it is the first time to solve the FDE and FTC problems for model-free MIMO discrete-time systems; second, the proposed FTC policy is simple to be implemented with reducing computational burden. Finally, two examples are used to demonstrate the effectiveness of the proposed FDE and FTC methods.