Condition Evaluation of Power Cables Based on BiLSTM with Siamese Attention

The fast and accurate state evaluation of cable plays an important role in the safe and reliable operation of power system. Thus a high-voltage cables condition evaluation model based on siamese-attention bidirectional long short-term memory (SA-BiLSTM) is proposed in this paper. First, the attention mechanism was used to optimize the BiLSTM network, which can complete parallel training while obtaining global information. then the optimized network was adopted to build a siamese neural network framework, realizing data classification by calculating the Euclidean distance between feature vectors. Finally, the model was trained and tested through the online monitoring data of high-voltage cables from engineering. The results showed that the accuracy rate, precision, recall rate, F1 score and Matthews correlation coefficient of the SA-BiLSTM model was increased at least by 7.92%, 11.79%, 11.41%, 12.04%, and 20.90%, respectively, compared with other models.