Fuzzy and Multilayer Perceptron for Evaluation of HV Bushings.pdf

Fuzzy and Multilayer Perceptron for Evaluation of HV Bushings Sizwe M. Dhlamini, Tshilidzi Marwala, and Thokozani Majozi Abstract—The work proposes the application of fuzzy set A. Background theory (FST) to diagnose the condition of high voltage bushings. There are 4 steps involved in fuzzy logic implementation, The diagnosis uses dissolved gas analysis (DGA) data from i.e. 1) Fuzzify inputs, 2) Select membership functions, 3) bushings based on IEC60599 and IEEE C57-104 criteria for oil Apply fuzzy operators, and finally 4) Defuzzify [1]. impregnated paper (OIP) bushings. FST and neural networks 1.1.1 Fuzzify inputs means: to identify the inputs or are compared in terms of accuracy and computational efficiency. Both FST and NN simulations were able to diagnose the attributes which describe the system. bushings condition with 10% error. By using fuzzy theory, the 1.1.2 Select membership functions means: to resolve all maintenance department can classify bushings and know the fuzzy statements (inputs) into a degree of membership extent of degradation in the component. between 0 and 1 for each attribute. 1.1.3 Apply fuzzy operators means: to AND or OR or I. INTRODUCTION NOT or ANY the inputs similarly to Boolean algebra. AND is THIS work presents fuzzy set theory (FST) used in the min fuzzy operat