Method for online SOH estimation of lithium-ion power batteries based on multi-factor capacity prediction empirical model

This article presents an online SOH estimation method for lithium-ion batteries using a multi-factor capacity prediction model. The model is trained using accelerated aging and basic performance tests, and the first-order RC parameter lines are used to identify the required OCV and R0. Historical data is fed into the model to obtain forward capacity predictions and correct the empirical model parameters. The Arrhenius model is used to simulate the power law relationship between capacity loss and cycle number, and the objective function is minimized to obtain optimal parameters. The model is represented as a discrete form of n and n-l cycles of capacity loss, with temperature, cycle number, and charging rate as parameters. Every n cycle, the capacity is calibrated, and the model parameters are corrected based on estimated error feedback for high-precision battery health status calculation.