Pattern recognition method for detecting fault in EHV transmission lines

Abstract

This paper presents a new approach for detecting and locating faults on an EHV transmission line using the method of pattern recognition. This technique of recognizing the waveform patterns can differentiate the fault condition with normal condition. A fault must be detected at its inception and issuing an output signal indicating this condition. Our approach is based on the fact that whenever fault occurs in any of the phases of transmission lines the impedance drops and current in the faulted phase rises. ANN backpropagation algorithm is used for training purpose, the learning process trained and tested a data set of several types of fault, the proposed method of detecting and locating faults gives great results which can support a new generation of very high speed protective relaying system.