A new method for fast state of charge estimation using retired battery parameters. (25th November 2022)

• Record Type: Journal Article
• Authors: Wang, Yuhang; Huang, Haihong; Wang, Haixin
• Is Part Of: Journal of energy storage. Volume 55:Part C(2022)
• Publication Date: 2022-11-25
• Subjects: Retired battery -- Equivalent circuit model -- Rapid -- State of charge -- Support vector machines; Energy storage -- Periodicals; Energy storage -- Research -- Periodicals; 621.3126
• DOI: 10.1016/j.est.2022.105621
• Languages: English
• ISSNs: 2352-152X

Abstract

Abstract: Model-based observers have been extensively studied for estimating battery healthy states. However, the parameters of the battery model are sensitive to the state of charge (SOC). Therefore, a fast SOC estimation method is needed. In this paper, a current pulse of 90 s is used to obtain the SOC characteristic parameters of the battery. Then, 2500 sets of SOC characteristic parameters at different SOC are used to train four different algorithms, among which the cubic support vector machine (Cubic SVM) method has the highest accuracy in estimating the SOC. The trained Cubic SVM model can quickly estimate the approximate SOC for 9 different capacity batteries with less than 5 % SOC estimation error. Meanwhile, the SOC-related parameters proposed in this paper can provide a reference for the fast sorting of retired batteries. Highlights: A fast SOC estimation method is proposed to improve the feasibility of using the battery model to estimate SOH. The relationship between some parameters and SOC of retired batteries has been described. The method proposed in this paper can estimate the approximate SOC in 90s.