Shaopeng Li , Hui Zhang , Daniela Anna Misul , Federico Miretti , Matteo Acquarone , Naikan Ding , Dingan Ni , Ninghao Hou , Yanjie He , Yijun Zhang , Yifan Sun
{"title":"基于时间卷积网络的电动汽车实际行驶工况下电池温度异常预警","authors":"Shaopeng Li , Hui Zhang , Daniela Anna Misul , Federico Miretti , Matteo Acquarone , Naikan Ding , Dingan Ni , Ninghao Hou , Yanjie He , Yijun Zhang , Yifan Sun","doi":"10.1016/j.etran.2025.100445","DOIUrl":null,"url":null,"abstract":"<div><div>For preventing thermal runaway accidents in electric vehicles (EVs), it is crucial to conduct early warning for temperature anomaly in battery pack. Based on data collected by a naturalistic driving experiment with 20 EVs, this study proposes a temporal convolutional network (TCN) algorithm for battery temperature anomaly prediction. Firstly, 40 features encompassing battery signals, thermal management state, ambient temperature, and driving condition are extracted from micro-segments. Then, the most effective input features are selected between the 40 features through maximum information coefficient (MIC) correlation analysis, and the principal component analysis (PCA). After obtaining the optimal hyperparameters, the TCN model is trained using the data from four EVs. The model's performance in predicting temperature is assessed over the data of the remaining 16 vehicles. The results demonstrate that the model achieves accurate prediction with the maximum and minimum mean relative error (MRE) of 0.0132 and 0.0072 across the 16 test vehicles. Moreover, the model proves to be robust against different testing seasons, SOCs, and traffic conditions. Compared to convolutional neural network (CNN), long short-term memory network (LSTM), and CNN-LSTM models with same hyperparameters, the developed TCN model consistently obtains the lowest MRE on both training and testing. For two kinds of scenarios where the probe temperature changes slowly and rapidly, the TCN model can predict an impending temperature anomaly up to 40 min in advance, and forecast the temperature anomaly within the future 8 min, respectively. Among the 16 vehicles, 81.25 % demonstrate a high prognosis accuracy, with an average F1 score of 0.951 across 10 of the vehicles. Thus, the proposed method can provide accurate battery temperature anomaly early warning for EVs under actual driving conditions.</div></div>","PeriodicalId":36355,"journal":{"name":"Etransportation","volume":"25 ","pages":"Article 100445"},"PeriodicalIF":15.0000,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Battery temperature anomaly early warning for electric vehicles under real driving conditions using a temporal convolutional network\",\"authors\":\"Shaopeng Li , Hui Zhang , Daniela Anna Misul , Federico Miretti , Matteo Acquarone , Naikan Ding , Dingan Ni , Ninghao Hou , Yanjie He , Yijun Zhang , Yifan Sun\",\"doi\":\"10.1016/j.etran.2025.100445\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>For preventing thermal runaway accidents in electric vehicles (EVs), it is crucial to conduct early warning for temperature anomaly in battery pack. Based on data collected by a naturalistic driving experiment with 20 EVs, this study proposes a temporal convolutional network (TCN) algorithm for battery temperature anomaly prediction. Firstly, 40 features encompassing battery signals, thermal management state, ambient temperature, and driving condition are extracted from micro-segments. Then, the most effective input features are selected between the 40 features through maximum information coefficient (MIC) correlation analysis, and the principal component analysis (PCA). After obtaining the optimal hyperparameters, the TCN model is trained using the data from four EVs. The model's performance in predicting temperature is assessed over the data of the remaining 16 vehicles. The results demonstrate that the model achieves accurate prediction with the maximum and minimum mean relative error (MRE) of 0.0132 and 0.0072 across the 16 test vehicles. Moreover, the model proves to be robust against different testing seasons, SOCs, and traffic conditions. Compared to convolutional neural network (CNN), long short-term memory network (LSTM), and CNN-LSTM models with same hyperparameters, the developed TCN model consistently obtains the lowest MRE on both training and testing. For two kinds of scenarios where the probe temperature changes slowly and rapidly, the TCN model can predict an impending temperature anomaly up to 40 min in advance, and forecast the temperature anomaly within the future 8 min, respectively. Among the 16 vehicles, 81.25 % demonstrate a high prognosis accuracy, with an average F1 score of 0.951 across 10 of the vehicles. Thus, the proposed method can provide accurate battery temperature anomaly early warning for EVs under actual driving conditions.</div></div>\",\"PeriodicalId\":36355,\"journal\":{\"name\":\"Etransportation\",\"volume\":\"25 \",\"pages\":\"Article 100445\"},\"PeriodicalIF\":15.0000,\"publicationDate\":\"2025-07-03\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Etransportation\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S2590116825000529\",\"RegionNum\":1,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"ENERGY & FUELS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Etransportation","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2590116825000529","RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENERGY & FUELS","Score":null,"Total":0}
Battery temperature anomaly early warning for electric vehicles under real driving conditions using a temporal convolutional network
For preventing thermal runaway accidents in electric vehicles (EVs), it is crucial to conduct early warning for temperature anomaly in battery pack. Based on data collected by a naturalistic driving experiment with 20 EVs, this study proposes a temporal convolutional network (TCN) algorithm for battery temperature anomaly prediction. Firstly, 40 features encompassing battery signals, thermal management state, ambient temperature, and driving condition are extracted from micro-segments. Then, the most effective input features are selected between the 40 features through maximum information coefficient (MIC) correlation analysis, and the principal component analysis (PCA). After obtaining the optimal hyperparameters, the TCN model is trained using the data from four EVs. The model's performance in predicting temperature is assessed over the data of the remaining 16 vehicles. The results demonstrate that the model achieves accurate prediction with the maximum and minimum mean relative error (MRE) of 0.0132 and 0.0072 across the 16 test vehicles. Moreover, the model proves to be robust against different testing seasons, SOCs, and traffic conditions. Compared to convolutional neural network (CNN), long short-term memory network (LSTM), and CNN-LSTM models with same hyperparameters, the developed TCN model consistently obtains the lowest MRE on both training and testing. For two kinds of scenarios where the probe temperature changes slowly and rapidly, the TCN model can predict an impending temperature anomaly up to 40 min in advance, and forecast the temperature anomaly within the future 8 min, respectively. Among the 16 vehicles, 81.25 % demonstrate a high prognosis accuracy, with an average F1 score of 0.951 across 10 of the vehicles. Thus, the proposed method can provide accurate battery temperature anomaly early warning for EVs under actual driving conditions.
期刊介绍:
eTransportation is a scholarly journal that aims to advance knowledge in the field of electric transportation. It focuses on all modes of transportation that utilize electricity as their primary source of energy, including electric vehicles, trains, ships, and aircraft. The journal covers all stages of research, development, and testing of new technologies, systems, and devices related to electrical transportation.
The journal welcomes the use of simulation and analysis tools at the system, transport, or device level. Its primary emphasis is on the study of the electrical and electronic aspects of transportation systems. However, it also considers research on mechanical parts or subsystems of vehicles if there is a clear interaction with electrical or electronic equipment.
Please note that this journal excludes other aspects such as sociological, political, regulatory, or environmental factors from its scope.