Fault analysis of analog electronic systems: Algorithms based on fuzzy sets

Three fundamental problems are involved in achieving effective automatic generation of fault isolation tests for analog electronic systems feature extraction, fault classification, and diagnosis. Focus on dealing with the fault isolation problem involves various kinds of uncertainties such as indeterminacy or vagueness. Such problems lend themselves very well to and in fact can be solved by adopting fuzzy set concepts. This line of research produced a modified fuzzy set technique applicable to automatic fault isolation. A base line automatic isolation system is used to deal with various kinds of uncertainties. A fuzzy automation model served as a point of departure for the base line system. Various fuzzy relations are used to select and update the parameters and structures of the system. Set of algorithms and new decision criteria are implemented easily and used for effective fault isolation. A fuzzy distance measure and a fuzzy entropy measure are used for decision making in the fault isolation algorithms.