Identification of electromagnetic transients in power transformer system using artificial neural network

This paper presents a novel technique for transient identification in power transformers based on the detection of the switching operation and fault generated high frequency signals using neural networks. A specially designed transient detector unit is first applied to capture the various transient signals, the captured signals are then used to train a neural network which is subsequently used to determine the source and nature of a transient. The simulation results show that the proposed technique is able to not only capture the high frequency current transient signals inside a transformer, but also accurately identify the source and nature of a transient