An adaptive observer for actuator and sensor fault diagnosis in linear time-varying systems

Fault detection and isolation is important and effective to improve the safety and reliability of engineering processes and systems. However, for systems with redundant sensors, it is important to detect faulty sensors as early as possible to reduce their effect on the performance of the control systems. In this paper, an observer for linear time-varying systems with both sensor and actuator faults is derived by extending the one that considers only actuator faults. Guidelines for selection the design parameters of the observer are discussed. The global exponential convergence is established for noise-free systems, whilst for noisy systems, the estimation errors are shown to be bounded and converged in the mean to zero if the noises are bounded and have zero means. The proposed technique is applied to detect both actuator and sensor faults in a satellite navigation system.