Scholarlink

Electric vehicles (EVs) are gaining popularity as a crucial component of smart mobility in smart city applications due to their significant role in lowering greenhouse gas emissions. Accurately determining the range of electric vehicles is one of the most difficult things. This might be a problem for riders who need to travel significant distances or have limited availability of charging outlets. In order to address this issue, the majority of manufacturers provide range forecasts that are derived from a combination of test data and real-world use scenarios. Hence, this research used a comparative analysis of artificial neural network (ANN) and convolutional neural network (CNN) models to forecast the range. Finally, the two models were compared and assessed using the mean square error (MSE). Based on the comparison, it can be inferred that the CNN model yields a more precise outcome, as shown by the lowest loss value of 3951.55.

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

Comparative Study of EV Range Prediction Using Deep Learning Algorithms

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Jigar Sarda,
  • Nilay Patel,
  • Rohan Vaghela,
  • Hirva Patel,
  • Akash Bhoi,
  • Arpita Patel,
  • Mohendra Roy

摘要

Electric vehicles (EVs) are gaining popularity as a crucial component of smart mobility in smart city applications due to their significant role in lowering greenhouse gas emissions. Accurately determining the range of electric vehicles is one of the most difficult things. This might be a problem for riders who need to travel significant distances or have limited availability of charging outlets. In order to address this issue, the majority of manufacturers provide range forecasts that are derived from a combination of test data and real-world use scenarios. Hence, this research used a comparative analysis of artificial neural network (ANN) and convolutional neural network (CNN) models to forecast the range. Finally, the two models were compared and assessed using the mean square error (MSE). Based on the comparison, it can be inferred that the CNN model yields a more precise outcome, as shown by the lowest loss value of 3951.55.