An ANFIS Based Study for High Impedance Faults in Distribution Systems

In this book, the application of Adaptive Neuro-Fuzzy Inference System (ANFIS) for protection of distribution systems is introduced. The ANFIS can be viewed as a fuzzy system, a neural network or fuzzy neural network. The objective of this research is firstly, to detect the High Impendence Fault occurrence, secondly to classify the fault type and deduce if it is single line to ground, double lines to ground, triple lines to ground or line to line. Finally, locating the fault will be carried out to estimate the fault distance and hence to eliminate the fault and decrease the outage time. The Fault Detectors, Classifiers and Locators are firstly trained using ANFIS technique and then they are tested in diversity of system conditions with respect to the fault types, distance, supply voltage, inception instant, resistance, and pre-fault conditions.