A remaining capacity estimation approach of lithium-ion batteries based on partial charging curve and health feature fusion. (November 2021)

• Record Type: Journal Article
• Authors: Fan, Lingfeng; Wang, Ping; Cheng, Ze
• Is Part Of: Journal of energy storage. Volume 43(2021)
• Publication Date: 2021-11
• Subjects: Lithium-ion battery -- Health feature -- State of health -- Gaussian process regression; Energy storage -- Periodicals; Energy storage -- Research -- Periodicals; 621.3126
• DOI: 10.1016/j.est.2021.103115
• Languages: English
• ISSNs: 2352-152X

Abstract

Highlights: Health features extracted from partial voltage, current and temperature curves. Feature optimization based on canonical correlation analysis. Remaining capacity estimation with gaussian process regression. Validations with battery data in various operating conditions. Abstract: Health condition monitoring of lithium-ion batteries plays a crucial role in guaranteeing the reliability and safety of energy storage system. However, it is difficult to directly measure state of health of cell, which is sensitive to the complex application scenarios and related to battery internal physicochemical characters. This paper proposes a remaining capacity prediction technique for lithium-ion batteries based on partial charging curve and health feature fusion. Here, health features are extracted from partial charging profile and the fused health feature is obtained via canonical correlation analysis, which is utilized to improve feature correlation and reduce data dimension. A battery aging model is established by Gaussian process regression and validation is conducted on six battery data sets from National Aeronautics and Space Administration (NASA) and University of Oxford. The results show that the proposed method can achieve reliable and accurate battery capacity estimation.