Remaining useful life Prediction for lithium-ion battery based on CEEMDAN and SVR

摘要

Estimation of lithium-ion battery remaining useful life (RUL) is the key to lithium-ion battery health. Achieving accurate and reliable remaining useful life prediction of lithium-ion batteries is very vital for the normal operation of the battery system. This paper proposes a lithium-ion battery RUL prediction method based on the combination of complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and support vector machine regression (SVR) which is multiple input and single output. First, a measurable health factor is extracted during the discharge process, and the correlation between health factor and capacity is analyzed by Pearson and Spearman methods. Then, the health factor is decomposed by CEEMDAN to obtain a series of relatively stable components. Finally, the health factor decomposed by CEEMDAN is used as the input of SVR prediction model, and the capacity is used as the output, to realize lithium-ion RUL prediction. Based on the lithium-ion battery degradation data set provided by NASA PCoE, the effectiveness of the proposed RUL prediction model is verified.