Lithium-ion battery state of health estimation with short-term current pulse test and support vector machine. (September 2018)

• Record Type: Journal Article
• Authors: Meng, Jinhao; Cai, Lei; Luo, Guangzhao; Stroe, Daniel-Ioan; Teodorescu, Remus
• Is Part Of: Microelectronics and reliability. Volume 88/90(2018)
• Page Start: 1216
• Page End: 1220
• Publication Date: 2018-09
• Subjects: State of health -- Lithium-ion battery -- Current pulse test -- Feature selection -- Support vector machine; Electronic apparatus and appliances -- Reliability -- Periodicals; Miniature electronic equipment -- Periodicals; Appareils électroniques -- Fiabilité -- Périodiques; Équipement électronique miniaturisé -- Périodiques; Electronic apparatus and appliances -- Reliability; Miniature electronic equipment; Periodicals; 621.3815
• DOI: 10.1016/j.microrel.2018.07.025
• Languages: English
• ISSNs: 0026-2714

Abstract

Abstract: State of Health (SOH) of Lithium-ion (Li-ion) battery plays a pivotal role in the reliability and safety of the Battery Energy Storage System (BESS) in the power system. Utilizing the features from the terminal voltage response of the Li-ion battery under current pulse test, a new method is proposed in this paper by using the Support Vector Machine (SVM) technique for accurately estimating the battery SOH. Since the terminal voltage measured at the same condition varies with the battery aging process, the features for SOH estimation are extracted from the voltage response under a specific current pulse test. The benefit of the proposed method is that the features come from the short-term test, which is much convenient to be obtained in real applications. After applying the short term current pulse test (few seconds), the keen points and the slopes in the voltage response curve are selected as the potential candidate features. In order to find the most effective feature for SOH estimation, all the possible combinations of the features are investigated and compared. Afterwards, SVM is able to establish the optimal SOH estimator on the basis of the optimal feature combination and the battery SOH. A LiFePO4 battery is tested in the test station for 37 weeks to verify the validation of the proposed method. Highlights: Features for SOH estimation are extracted from the short-term current pulse test. Optimal feature is selected from all the candidate features. Support vector machine is used to establish the SOH estimator. The proposed method is validated on a LiFePO4 battery with 37 weeks' test.