Transformer Oil Temperature Prediction Based on WOA-CNN-BiLSTM Model with Dual Attention Mechanism
The current transformer oil temperature prediction model is difficult to make full use of the large amount of time series data,a dual attention optimization WOA-CNN-BiLSTM transformer oil temperature prediction model is proposed.The model takes the transformer historical oil temperature data as input,upgrades the input data through the sliding window method,mines the feature relations by combining the Channel Attention Mechanism(SENet),the squeeze excitation networks(SENet),with the convolutional neu-ral network(CNN),and Bidirectional long and short-term memory network(BiLSTM)is used to further excavate the temporal features.Then,Time Attention Mechanism is added for weight allocation,and the established model is used for hyperparameter opti-mization using whale algorithm to realize the transformer oil temperature prediction.Fi-nally,the measured data are used for case analysis to verify the superiority of this method.
万方数据
维普
