A Strategy for Predicting Residual Service Life of Batteries Based on Impedance Spectrum

Abstract With the wide application of lithium batteries in electric vehicles, energy storage and other fields, accurate life prediction ion becomes particularly important. As a fast non-destructive electrochemical technology, impedance spectroscopy can reflect internal characteristics Li-ion batteries, it is widely used battery management. A method based on health characteristic parameters extracted from spectrum proposed this paper to predict life. Firstly, equivalent circuit model fit screen out quantity which aging state battery. Then, BP neural network was establish relationship between two. Finally, particle filter algorithm remaining service At end paper, measured results are compared with predicted verify effectiveness method.