Scholarlink

EVs (Electric vehicles) have become commonplace, and their appeal has risen owing to air pollution plus higher fuel prices, thereby becoming a popular green mode of transportation. EVs aid a transformative shift in sustainable transportation, offering advantages such as reduced emissions, lower operational costs, and enhanced driving efficiency compared to internal combustion engine (ICE) and hybrid electric vehicles (HEVs). This paper presents a comprehensive comparative analysis of ICE, HEV, and EV systems, evaluating their performance, energy efficiency, and environmental impact. In electric vehicles, the battery stores electrical power. The study highlights lithium-ion batteries (Li-ion) as the most suitable energy storage solution for EVs. The DC-to-DC converter couples the inverter to the battery via direct current (DC). A motor transmits the motion of the vehicle. As a consequence, this cutting-edge resource includes comprehensive information on power electronics converters and battery management systems (BMS), along with motors. This research proposes a closed-loop machine learning framework for early prediction of battery failure by analyzing voltage drop profiles and integrating probabilistic models with real-world battery data. The developed BMS incorporates cloud-based digital twins and longitudinal health monitoring to enhance predictive accuracy and prevent false alarms. Various machine learning techniques are suggested to assure the safety of automobile batteries by improving the classification accuracy from the baseline model, thereby necessitating the requirements for robust and accurate battery failure prediction for specifically tailored cloud-based artificial intelligence solutions. Lastly, the research looks into the battery’s thermal runaway mechanism based on the noted battery operating data, as well as pinpoints the imperative failure signal time points that confirm the commencement and succession of thermal runaway in the battery pack. The findings contribute to advancing sustainable transportation by integrating empirical, physical, and AI-driven approaches for next-generation EV systems.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Extensive Analysis of Electric Vehicles: Battery Management Systems, Machine Learning Strategies to Prevent Thermal Runaway for Improved Safety and Performance

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Dinesh Kumar,
  • Mohit Kumar Sharma

摘要

EVs (Electric vehicles) have become commonplace, and their appeal has risen owing to air pollution plus higher fuel prices, thereby becoming a popular green mode of transportation. EVs aid a transformative shift in sustainable transportation, offering advantages such as reduced emissions, lower operational costs, and enhanced driving efficiency compared to internal combustion engine (ICE) and hybrid electric vehicles (HEVs). This paper presents a comprehensive comparative analysis of ICE, HEV, and EV systems, evaluating their performance, energy efficiency, and environmental impact. In electric vehicles, the battery stores electrical power. The study highlights lithium-ion batteries (Li-ion) as the most suitable energy storage solution for EVs. The DC-to-DC converter couples the inverter to the battery via direct current (DC). A motor transmits the motion of the vehicle. As a consequence, this cutting-edge resource includes comprehensive information on power electronics converters and battery management systems (BMS), along with motors. This research proposes a closed-loop machine learning framework for early prediction of battery failure by analyzing voltage drop profiles and integrating probabilistic models with real-world battery data. The developed BMS incorporates cloud-based digital twins and longitudinal health monitoring to enhance predictive accuracy and prevent false alarms. Various machine learning techniques are suggested to assure the safety of automobile batteries by improving the classification accuracy from the baseline model, thereby necessitating the requirements for robust and accurate battery failure prediction for specifically tailored cloud-based artificial intelligence solutions. Lastly, the research looks into the battery’s thermal runaway mechanism based on the noted battery operating data, as well as pinpoints the imperative failure signal time points that confirm the commencement and succession of thermal runaway in the battery pack. The findings contribute to advancing sustainable transportation by integrating empirical, physical, and AI-driven approaches for next-generation EV systems.