LAUSR

Abstract Accurate real-time State-of-Health (SoH) estimation is important in ensuring battery safety and promoting efficiency. The open circuit voltage (OCV), relative to State-of-Charge (SoC), changes as a battery degrades. In this paper, we propose using an extended Kalman filter-recursive least squares parameter identification method based on the second-order RC equivalent circuit model. SoC estimation, RC model parameter identification and OCV-SoC relationship revision can be conducted simultaneously at three timescales to improve the identification accuracy. Then, by analyzing the relationship between battery SoH and the model parameters, we establish a support vector regression algorithm for online SoH estimation. Finally, the dynamic stress test and the federal urban driving schedule test are adopted and used as practical simulations to verify the feasibility of the method under dynamic conditions. The results show that the proposed method achieves highly accurate estimations and has strong robustness under both static and dynamic conditions.