A Transformer Vibration Signal Separation Method Based on BP Neural Network

The power transformer is one of the most important power equipment in the power system. Its operation reliability is related to the safe and stable operation of the power system. Therefore, the transformer fault diagnosis has been paid much attention by the researchers. The mechanical fault diagnosis method based on the vibration signal of the tank surface has been widely studied because the measurement system has no direct electrical connection with the transformer and has a strong anti-interference ability. The traditional vibration signal analysis method generally analyzes the mixed signal on the surface of the transformer tank, and can't effectively evaluate the mechanical state of the winding and the core. Therefore, it is important to carry out the research on the vibration signal separation technology of transformer. In this paper, a vibration signal separation technique for transformer oil tank surface based on back-propagation (BP) neural network is proposed. The average value of the waveform similarity coefficient of the core vibration signal is 0.813, the average value of the waveform similarity coefficient of the winding vibration signal is 0.834. It provides an important technical method for the effective evaluation of the mechanical state of the winding and the core.