Capacity prediction of lithium-ion batteries based on ensemble empirical mode decomposition and hybrid machine learning

Abstract

Considering the influence of capacity regeneration on the prediction accuracy of the remaining useful life (RUL) of lithium-ion batteries (LIB), a multi-stage capacity prediction method based on ensemble empirical mode decomposition (EEMD) and hybrid machine learning is proposed. Firstly, the aging data of LIB is decomposed into residual sequence (degradation trends) and intrinsic mode function (IMF) by the EEMD algorithm. Next, the long short-term neural network model with Bayesian optimization and the support vector regression model optimized by the improved whale algorithm were used to model and predict the decomposed IMF components and residual sequences. The predicted residual and IMF data are integrated to calculate the future life aging trajectory of LIB and further extrapolate to obtain the predicted RUL value. Finally, different battery aging data are used to verify the proposed method, and the offline prediction results show that the proposed method has high prediction accuracy and generalization adaptability.