Prediction of Remaining Life of Relay Protection Equipment Using Transformer-Holt Model

Accurately predicting the remaining service life of relay protection equipment is crucial for maintaining the stability and safety of power systems. This study proposes a prediction method, based on the Transformer-Holt model, to improve the accuracy and generalization ability of predicting the remaining life of relay protection equipment. This method combines the deep learning ability of the Transformer model and the time-series smoothing processing function of the Holt model to develop a new type of prediction model. The experimental results showed that on the Commercial Modular Aero-Propulsion System Simulation (CMAPSS) dataset, the prediction accuracy of this model reached 0.98, the prediction delay score was 0.90, and the generalization ability score was 0.95, all of which were significantly better than the comparison models. In the actual prediction scenarios, the evaluation parameters of mean square error, root mean square error, and mean absolute error of the model were 0.05, 0.21 and 0.15 respectively, demonstrating excellent performance. The proposed Transformer-Holt model provides a new technical approach to the predictive maintenance of power systems.