A Novel Algorithm for Discrimination of the Magnetizing Inrush Current and Internal Fault Current of a Transformer Using Teager Energy Operator and Artificial Neural Network

The transformer is a very important component of the power system. Hence, it needs to protect the transformer from various faults occurring in the system with reliable and sensitive protective system. In the literature, there are many methods available to protect the transformer from abnormalities. Differential protection is the most commonly used protection method. But, it is observed that the conventional differential protection scheme mal-operates because of magnetizing inrush current. This paper presents a Teager energy operator (TEO) and artificial neural network (ANN)-based novel approach for differentiating between the magnetizing inrush current and the internal fault current of a transformer, which prevents the differential relay from malfunctioning and hence increases the system reliability. In this paper, the differential current of the transformer captured from the experimentation is used to calculate the TEO. The values of TEO and the threshold value are compared in order to detect the occurrence of abnormality. Once the abnormality is detected a detection flag is set, and the statistical parameters are calculated as a feature from the differential current for the quarter cycle from the instant of detection. A feature vector is created from the extracted features to train and test the ANN classifier. This paper also proposes an additional algorithm based on the duration of the detection flag. The proposed algorithms are tested using the experimental data. The results show that the suggested algorithms are capable of truthfully differentiating between the transformer’s internal fault current and inrush current.