State of health estimation of large-cycle lithium-ion batteries based on error compensation of autoregressive model. (15th August 2022)

• Record Type: Journal Article
• Authors: Feng, Hailin; Yan, Huimin
• Is Part Of: Journal of energy storage. Volume 52:Part B(2022)
• Publication Date: 2022-08-15
• Subjects: Lithium-ion batteries -- State of health -- Autoregressive model -- Relevance vector machine -- Error compensation; Energy storage -- Periodicals; Energy storage -- Research -- Periodicals; 621.3126
• DOI: 10.1016/j.est.2022.104869
• Languages: English
• ISSNs: 2352-152X

Abstract

Abstract: Accurate assessment of Lithium-ion battery state of health (SOH) is the core issue of Lithium-ion battery safety management. A novel SOH estimation model for large-cycle lithium-ion batteries with error compensation mechanism—combining autoregressive model, equal voltage discharge time and relevance vector machine (AR-EVDTRVM) model is proposed in this paper. Firstly, the SOH prediction framework of large-cycle Lithium-ion batteries based on the autoregressive (AR) model is established. Then, using the relevance vector machine (RVM) algorithm, an error compensation mechanism characterized by the equal voltage discharge time (EVDT) of Lithium-ion batteries is proposed. Finally, the experimental verification of AR-EVDTRVM model is carried out on the large-cycle and small-cycle lithium battery datasets, respectively. Compared with other methods, results show that the AR-EVDTRVM model has high estimation accuracy and strong robustness. Highlights: A model for estimating the state of health of a large-cycle lithium-ion battery is established. For problems with large prediction errors in the Autoregressive model, an error compensation mechanism is used. The error compensation mechanism is based on the discharge index and the Relevance vector machine algorithm.