RECOGNITION OF POWER SYSTEM TRANSIENTS AND FAULT ANALYSIS BASED DISCRET WAVELET TRANSFORM AND NEURAL NETWORE K

摘要

Now days, because of the great extent of capitalrninvestment to generate electrical power supply, powerrnsupply has become as an article of trade. It is extremelyrnsignificant to serve the power in reliable and stable needrnto ensure customers requirement. To attain stability and tornensure power quality performance it is very important tornidentify power system faults. High impedance fault (HIF)rnis one of the various fault causes; distinctiverncharacteristics of this type of faults are asymmetry andrnnonlinearly behavior. Arc which is usually associatedrnwith these kinds of faults is considered as a source ofrnhuman life risks and fire hazardous. Hence, detection andrnprotection of such faults still remain a topic of study. Inrnthis paper PSCAD/EMTDC simulation, software is usedrnto simulate the high-voltage transmission system andrnmodeling of HIF. MATLAB software is used for;rnadvanced signal processing tools such as discrete waveletrntransform (DWT) which is used as feature extraction tornget an useful information from faulted signal, backrnpropagation neural network (BP-NN) for detection ofrnhigh impedance fault.