Data-Driven Output-Feedback Fault-Tolerant Tracking Control Method and Its Application to a DC Servo System

The paper is concerned with the problem of data-driven output-feedback fault-tolerant tracking control design for unknown discrete-time linear systems with stochastic measurement and process noise. With the aid of a proposed input-output data-based online iterative identification algorithm, this study presents a fault detection strategy, an approximate dynamic programming approach to performance optimization, and a method for the achievement of tracking control via a pre-filter. The resulting fault-tolerant control (FTC) scheme is applied to a practical DC servo motor system with load fluctuations and nonlinearity. Finally, the experimental test demonstrates that the proposed FTC strategy (including a methodology for compensating load variations and nonlinearity) has better applicability and practicability than the one given by the previous work, where a load needs to be held constant and it is required that the motor run in a nearly linear small working region.