State of Health Estimation for Lithium-Ion Battery based on data-driven dual Kalman filtering

This article focuses on the actual operation data of pure electric vehicles in Qingdao in the past two years, and proposes a dual Kalman filtering algorithm to accurately monitor battery health, solving the problems of direct measurement method being unable to monitor in real time and model method being greatly affected by parameters. The SOH results calculated using Matlab indicate that during the testing period, the battery health status decreased slightly due to low temperatures in winter, but remained at around 80% overall. The mean square error of the dual Kalman model gradually decreases over time, with a minimum error of 0.0017863, demonstrating the reliability and effectiveness of the model in practical applications. Key Words: Data driven, real vehicle operation data, dual Kalman filtering, power battery SOH monitoring