State-of-Health Estimation Based on Differential Temperature for Lithium Ion Batteries

• Published in: IEEE transactions on power electronics Vol. 35; no. 10; pp. 10363 - 10373
• Main Authors: Tian, Jinpeng, Xiong, Rui, Shen, Weixiang
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.10.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Aging; Batteries; Battery ageing; battery management; Degradation; differential temperature; Differential thermal analysis; Electric potential; Estimation; Ions; Lithium; lithium ion battery; Lithium-ion batteries; Rechargeable batteries; Regression analysis; Robustness (mathematics); Sampling; state-of-health; Support vector machines; Temperature measurement; Voltage
• ISSN: 0885-8993; 1941-0107
• DOI: 10.1109/TPEL.2020.2978493

State-of-health (SOH) estimation is necessary for lithium ion batteries due to ineluctable battery ageing. Existing SOH estimation methods mainly focus on voltage characteristics without considering temperature variation in the process of health degradation. In this article, we propose a novel SOH estimation method based on battery surface temperature. The differential temperature curves during constant charging are analyzed and found to be strongly related to SOH. Part of the differential temperature curves in a voltage range is adopted to establish a relationship with SOH using support vector regression. The influence of battery discrepancy, voltage range, and sampling step are systematically discussed and the best combination of voltage range and sampling step is determined using leave-one-out validation. The proposed method is then validated and compared with an incremental capacity analysis (ICA)-based SOH estimation method using the Oxford and NASA datasets, which were collected from different cells under different conditions, respectively. The results show that the proposed method is capable of estimating SOH with the root-mean-square error less than 3.62% and 2.49%, respectively. In addition, the proposed method can improve the overall SOH estimation accuracy and robustness by combining with the ICA-based method with little computational burden.