Estimation Method of Lithium Battery Health State Based on Space Detection

Research on the health status detection of lithium batteries mostly focuses on DC or AC detection methods excited by current or voltage. Such methods may need to be based on a large amount of charge and discharge data. Or it may be based on the analysis technology of electrochemical impedance spectroscopy (EIS) provided by precision detection equipment, which consumes more time and energy, and has higher detection cost and computing power cost. Based on the above situation, this paper proposes a spatial detection method based on electric field and magnetic field. This method realizes the health status prediction of lithium batteries by detecting the spatial electric field and magnetic field, avoiding the above problems. This paper establishes a simulation experiment platform for spatial detection of lithium batteries based on COMSOL Multiphysics software. The detection experiments verify the feasibility of applying spatial detection technology to health status prediction. In this paper, the random forest algorithm is adopted to establish a prediction model for the health status of lithium batteries. Through comparative experiments of modeling with different feature combinations, it is verified that features such as the equivalent capacitance density Cd, equivalent inductance density Ld, and equivalent characteristic impedance density Zd extracted based on the detection data of electric fields and magnetic fields can effectively improve the prediction accuracy of the model. The SOH prediction error of the random sample is 0.35%, which is 1.4% lower than that of the SOH prediction result modeled by the static port voltage in the fixed SOC state.