Discrimination of insulation defects using a neural network

신경회로망을 이용한 절연 결함의 판별

This paper describes the method of diagnosing the degradation by void defects of insulator inside in operation. Needle-shape void specimens, made from LDPE, were used to generate an electrical tree under ac voltage. The method uses a neural network system with input signal of AE patterns. AE pattern consists of the pulse count and average amplitude according to the phase angle. After the learning process was over, unknown emission patterns were put into the network. It was shown that the network discriminates the void deflects well. The effectiveness of the neural network system for partial discharge recognition was shown.