Remaining Useful Life Prediction of Li-Ion Batteries Based on GRU-Transformer

Accurate prediction of the Remaining Useful Life (RUL) of batteries is crucial for battery health management and predictive maintenance. This paper presents a novel hybrid Gate Recurrent Unit (GRU)-Transformer neural network approach to improve the accuracy of battery RUL prediction. The method employs a GRU network to extract local temporal features and a Transformer network to capture the complex dynamic changes in the battery degradation process. Furthermore, to optimize model performance, the Bayesian optimization algorithm is utilized to optimize the hyperparameters of the GRU-Transformer model. Experimental results demonstrate that, compared to methods that using the GRU or Transformer model individually, the proposed method achieves better performance on the battery RUL prediction task, with an MSE of 0.0014 and an R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of 95.34%.