Fault Detection and Diagnosis
In this chapter, we start with the formal definition and formulation of the fault detection and diagnosis problem in nonlinear systems. Then, desired attributes of a fault diagnosis system and the rationale behind each attribute are discussed. A comprehensive survey and analysis of the literature on model-based and computational intelligence (CI)-based approaches to fault diagnosis is then presented with individual emphasis on the tasks of detection, isolation and identification. A number of well-known methodologies within each approach are further demonstrated, and their respective advantages and disadvantages are highlighted. Finally, the issue of robustness in fault diagnosis is introduced and briefly discussed.