锂(药物)
离子
材料科学
计算机科学
化学
医学
有机化学
内分泌学
作者
Fengfei Wang,Shengjin Tang,Xiao Han,Yu Chen,Xiaoyan Sun,Languang Lu,Minggao Ouyang
出处
期刊:Energy
[Elsevier]
日期:2024-02-01
卷期号:: 130743-130743
标识
DOI:10.1016/j.energy.2024.130743
摘要
Accurately predicting the health status of batteries through easily available data is crucial for the battery management system (BMS) in electric vehicles. This paper utilizes the battery aging information obtained from the partial charging process as the input to accurately predict the battery charging process based on the bi-directional long short-term memory (Bi-LSTM) network. Firstly, in order to obtain a precise incremental capacity (IC) curve containing battery aging information, a generalized derivative method based on the back propagation (BP) neural network for discrete data is proposed. Then, using this derivative method, the interrelated three-dimensional features, namely charged capacity (Q), voltage (V), and the dQ/dV values, are identified from the IC curve containing a complete peak with an interval length of 100 mV. Subsequently, a Bi-LSTM network is constructed to learn the correlation between the three-dimensional features and charging capacity, and predict the charging process within the interval of 3.8V–4.2V. Finally, the Oxford battery degradation dataset is used for verification. The results show that the battery aging information extracted during the partial charging process is closely related to battery capacity degradation, and the proposed capacity prediction method with fusing aging information can accurately predict the charging process of lithium-ion batteries.
科研通智能强力驱动
Strongly Powered by AbleSci AI