A fuzzy chip-based real-time fault classifier in a power controller

A fuzzy chip-based electrical power faults classifier is presented in this paper. The system, which utilizes a fuzzy chip designed for the fuzzy rule base inference, detects the faults in the electrical power system in real time and activates the circuit control unit to take the appropriate actions. A set of features are extracted, and two sets of fuzzy inference rules are used to classify faults based on those features. The membership functions for all fuzzy variables are trained based on a supervised learning algorithm. Features extracted from structure properties of the patterns enable the classifier to rapidly detect the faults appearing in electrical power within 50 μs. The fuzzy chip, in this fault classifier, provides speed and cost improvement over the existing general-purpose microprocessor technologies,