WAVELET TRANSFORM BASED FAULT DETECTION AND CLASSIFICATION IN POWER TRANSMISSION LINES

Abstract

Transmission Line faults are inevitable and a cause for power system instability. So to protect the network from severe consequences one need to identify the faulty section and take stringent and suitable protection to forefend the perilous outcome. To detect and identify the instance of fault and its types is one difficult task. The necessary job can be performed by the help of Classification algorithms and different evolutionary techniques. In this paper the wavelet transform and approximation coefficient energies is used to analyse the transients which caused during the fault and the performance and behaviour of the power system is studied during such adverse conditions. Artificial Neural network based classification using single layer feed forward network model is used to classify the faults using conventional gradient based algorithm and Extreme Learning Machine which has extremely fast learning capability is also used to train under various circumstances which was studied. The objectives of this work include: Study of different types of Faults. To propose a novel detection algorithm. Simulate different faults in PSCAD environment. Implement the proposed scheme on MATLAB. Collect the features for classification Algorithm. Classification of fault using ANN and ELM.