A systematic model-based degradation behavior recognition and health monitoring method for lithium-ion batteries. (1st December 2017)

• Record Type: Journal Article
• Authors: Xiong, Rui; Tian, Jinpeng; Mu, Hao; Wang, Chun
• Is Part Of: Applied energy. Volume 207(2017)
• Page Start: 372
• Page End: 383
• Publication Date: 2017-12-01
• Subjects: Lithium-ion battery -- Degradation recognition -- Remaining capacity -- Polarization resistance -- Electrochemical impedance spectroscopy; Power (Mechanics) -- Periodicals; Energy conservation -- Periodicals; Energy conversion -- Periodicals; 621.042
• DOI: 10.1016/j.apenergy.2017.05.124
• Languages: English
• ISSNs: 0306-2619

Abstract

Highlights: A simplified impedance model was established from the EIS test. Online parameters identification method for simplified impedance model was proposed. Differences of model parameters identified by time-frequency domain were analyzed. SEI resistance was confirmed as the most sensitive parameter to indicate degradation. A novel capacity estimation method with the SEI resistance has been proposed. Abstract: Degradation is a complex and intricate process which relates strongly to the state of health (SoH) of a lithium-ion battery. Due to the ambiguous mechanism and sensitivity to the objective factors of lithium-ion batteries, it is difficult to recognize the degradation state and monitor the SoH of a battery. A recognition method for the degradation state to estimate the remaining capacity online has been presented. First, through the analysis of the results of electrochemical impedance spectroscopy (EIS) tests at different SoHs, the degradation level can be detected by the EIS measurement. Second, according to the fractional order theory, an online parameter identification approach with the fractional order impedance model has been proposed for the degradation analysis. Third, the correlation between variation of parameters and degradation level is discussed and the SEI (Solid Electrolyte Interphase) resistance is extracted to predict the remaining capacity by selecting an appropriate fitting function. Finally, the effectiveness of the presented method is validated by the test data, and the estimation error of the remaining capacity can be guaranteed within 3%.