Forecasting State-of-Health of Battery Based on Combined EMD-CNN-BiLSTM-SM Method

Battery energy storage systems (BESSs), especially Lithium-ion batteries, are widely applied in power systems due to their numerous advantages. However, the battery will be unstable, leading to failure in the system operation or causing safety accidents when the battery reaches its end-of-life (EOL). Therefore, to accurately predict the State of health (SOH) and the remaining useful life (RUL), a novel prediction method is proposed in this paper based on long short-term memory (LSTM), convolutional neural network (CNN), and self-attention mechanism (SM). The degradation data of the batteries are first processed by the Empirical Mode Decomposition (EMD), then the data is put into the proposed model, and the obtained prediction results are compared with other models for validation. Different batteries from two datasets are utilized in this work to determine the robustness of the method. The errors, including Root Mean Square Error (RMSE), and Mean Absolute Error (MAE) are also calculated to verify the accuracy, and the RMSE and MAE of the proposed method are about 32.6% and 22.7% compared to other methods in all cases, respectively.