储能科学与技术2025，Vol.14Issue(1)：380-387,8.DOI:10.19799/j.cnki.2095-4239.2024.0571
基于EEMD-GRU-NN锂离子电池表面温度预测方法研究
Lithium-ion batteries surface temperature prediction toward EEMD-GRU-NN method
摘要
Abstract
As global demand for sustainable energy increases,ensuring the safety of energy storage batteries has become crucial.Accurate prediction of battery temperature is essential for preventing overheating and reducing the risk of battery failure,fire,or explosion due to high temperatures,thereby improving device safety.This study introduces a combined prediction approach based on ensemble empirical mode decomposition,gated recurrent units,and a basic neural network(NN).Initially,lithium battery temperature data was decomposed into periodic and trend components,which serve as target values for offline supervised learning training.Next,suitable feature parameters based on the temperature characteristics of the battery were selected as input features for the model to create a real-time online prediction model.Finally,the outputs of the two models were superimposed to obtain the final prediction result.We demonstrated the accuracy of the proposed method by comparing it with common NN models.Experimental results indicate that under normal temperature conditions,the proposed method outperforms traditional models in all evaluation metrics,achieving a root mean square error of 0.10℃,an average absolute error of 0.075℃,and a maximum error of 0.34℃.Although the prediction capability of the model decreases under extreme conditions,the error remains within a reasonable range,confirming the robustness of the model under extreme conditions.关键词
锂离子电池/温度预测/集合经验模态分解/门控循环单元Key words
lithium-ion battery/temperature prediction/ensemble empirical mode decomposition/gated recurrent unit分类
信息技术与安全科学引用本文复制引用
叶石丰,洪朝锋,綦晓,吴伟雄,谭子健,周奇,张兆阳..基于EEMD-GRU-NN锂离子电池表面温度预测方法研究[J].储能科学与技术,2025,14(1):380-387,8.基金项目
南方电网科技项目(GDKIXM20230246(030100KC23020017)),国家自然科学基金(52106244),广东省基础与应用基础研究基金(2022A1515011936). （GDKIXM20230246(030100KC23020017）
