An indirect RUL prognosis for lithium-ion battery under vibration stress using Elman neural network

• Published in: International journal of hydrogen energy Vol. 44; no. 23; pp. 12270 - 12276
• Main Authors: Li, Wenhua, Jiao, Zhipeng, Du, Le, Fan, Wenyi, Zhu, Yazun
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 03.05.2019
• Subjects: Elman neural network; Health indicator (HI); Lithium-ion battery; Long short-term memory (LSTM); Remaining useful life (RUL); Vibration stress; Lithium-ion battery; Remaining useful life (RUL); Health indicator (HI); Vibration stress; Long short-term memory (LSTM); Elman neural network
• ISSN: 0360-3199; 1879-3487
• DOI: 10.1016/j.ijhydene.2019.03.101

Remaining useful life (RUL) prognosis of lithium-ion battery can appraise the battery reliability to determine the advent of failure and mitigate risk. To acquire measurement data at similar working conditions as electrical vehicles (EVs), this paper mainly conducted the experiment about battery charging and discharging under vibration stress. Indirect health indicator (HI) was extracted from the time of equal discharge voltage from the upper to the lower, and the battery capacity proved to be estimated by the adopted indirect HI through grey relational analysis. Then, the RUL prognosis model based on Elman neural network was established. Finally, the feasibility of this RUL prognosis model based on Elman neural networks in an application in predicting RUL of battery under vibration stress was verified. •The experiments under vibration stress was conducted.•Indirect HI was extracted from voltage features of battery under vibration stress.•The indirect RUL prognosis based on Elman neural network was established.•The feasibility of this prognosis model in an application was verified.