{"title":"Convolutional Neural Network-Based Classification of Lithium-Ion Battery CAN Messages","authors":"Tero Niemi;Tomi Pitkäaho;Juha Röning","doi":"10.1109/OJVT.2025.3541382","DOIUrl":null,"url":null,"abstract":"The lithium-ion battery Controller Area Network (CAN) messages are essential to battery monitoring, recycling, and second-life applications. However, the proprietary nature of database connection (DBC) files and the diversity of CAN message encodings across manufacturers pose significant challenges. This study proposes a convolutional neural network (CNN) based approach to classify battery-related CAN messages without reliance on proprietary DBC files. By analyzing data from four manufacturers and categorizing messages into three key groups—voltage and current, temperature and State of Charge (SoC), and configuration or other battery parameters, the CNN achieved an accuracy of 94.87% on unseen data. The model demonstrated robust performance, effectively generalizing across diverse CAN message formats. Practical validation confirmed the model's ability to identify key battery metrics reliably. This publication highlights the potential of deep learning to address proprietary data barriers, facilitating accessible and scalable battery monitoring and health assessment approaches. The findings contribute to advancing sustainable battery management practices, particularly for companies focusing on battery recycling and second-life applications, and pave the way for further research on leveraging temporal and expanded datasets to enhance classification accuracy and scope.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"6 ","pages":"790-800"},"PeriodicalIF":5.3000,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10884024","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Open Journal of Vehicular Technology","FirstCategoryId":"1085","ListUrlMain":"https://ieeexplore.ieee.org/document/10884024/","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0
Abstract
The lithium-ion battery Controller Area Network (CAN) messages are essential to battery monitoring, recycling, and second-life applications. However, the proprietary nature of database connection (DBC) files and the diversity of CAN message encodings across manufacturers pose significant challenges. This study proposes a convolutional neural network (CNN) based approach to classify battery-related CAN messages without reliance on proprietary DBC files. By analyzing data from four manufacturers and categorizing messages into three key groups—voltage and current, temperature and State of Charge (SoC), and configuration or other battery parameters, the CNN achieved an accuracy of 94.87% on unseen data. The model demonstrated robust performance, effectively generalizing across diverse CAN message formats. Practical validation confirmed the model's ability to identify key battery metrics reliably. This publication highlights the potential of deep learning to address proprietary data barriers, facilitating accessible and scalable battery monitoring and health assessment approaches. The findings contribute to advancing sustainable battery management practices, particularly for companies focusing on battery recycling and second-life applications, and pave the way for further research on leveraging temporal and expanded datasets to enhance classification accuracy and scope.