Life Prediction of Lithium-Ion Batteries Based on Fusion Techniques

Prediction of the remaining service life of lithiumion batteries can detect battery aging or abnormalities in time, which is conducive to the adoption of appropriate safety measures to avoid safety hazards of batteries in the process of use, and is of great signiffcance for electric vehicles, portable electronic devices and other products that require long-term stable use. Aiming at the current lithium-electronic battery life prediction method can not simultaneously extract lithium-ion battery data sequence features and modal features of the problem, a fusion technology of lithium-ion battery life prediction method is proposed. The method fuses the model-driven and data-driven methods by wavelet transform, and wavelet decomposes the signal to obtain the frequency domain information of the signal on different scales, which overcomes the limitations of the single data-driven method. The experimental results show that the research on life prediction of lithium-ion batteries based on the fusion technique has achieved signiffcant improvement compared to the existing state-of-the-art technique (SOTA).