A Transformer Vibration Amplitude Prediction Method Via Fusion Of Multi-Signals

With the increase of the electrical equipment in the power system, the operation state of power transformer will further influence the security and stability of the power system. Generally, transformer surface vibration amplitude can indicate the operation status of transformer. So, a transformer surface vibration amplitude prediction model, which considered multi-signals, is proposed in this paper. The proposed method is based on CNN-LSTM model, where CNN is used to complete the feature extraction, and LSTM network is used to predict the transformer surface vibration amplitude through the extracted features. And the modified CNN-LSTM method can handle the multi-signals very well and increases the prediction performance compared with existing LSTM and CNN-LSTM. Followed by analysis, the experimental evaluation validates the claimed advantages of the proposed model.