Extraction of features to discriminate and detect transformer inrush current and faulty condition using ANN

The inrush current occurs when transformers are initially energized at no load, which lead to mal-operations of protective relays and issues of power quality. However, it is very necessary to discriminate between inrush condition and faulty condition for modern protection of power transformer. This paper presents an alternative approach of using artificial neural network for discrimination and detection of transformer inrush current and faulty condition by extracting features. In this paper, second harmonics ratio (SHR), peak value of inrush current and mean deviation are the features extracted which are used as input to ANN for discrimination between inrush and faulty condition.