Research on the Method of Remaining Useful Life Prediction of Lithium-ion Battery Based on LSTM

Lithium-ion battery is the most widely used storage battery nowadays. Accurate estimating the degradation of lithium-ion batteries and predicting its remaining useful life is critical for operational maintenance. This paper presents a prediction model for the remaining useful life of the battery based on Long Short-Term Memory neural networks. The battery capacity is used as a indicator of lithium-ion battery degradation, and data-driven capacity is used as forecasting methods. The prediction of lithium-ion battery life was realized after the degradation characteristics was extracted and the neural network structure and the relevant parameters was optimized. We complete the verification experiment based on the data set of 18650 lithium batteries provided by NASA Ames research center. The results show that the prediction model of remaining useful life fit the real lithium-ion battery characters better. It is able to make accurate predictions of the residual life of lithium-ion batteries and have good ability to use and save test time and cost for actual maintenance..