Enhancing Electric Vehicle Remaining Range Prediction Through Machine Learning

Many automakers are announcing electric vehicle (EV) models in response to environmental regulations. However, charging times still exceed those of traditional internal combustion engine vehicles. Moreover, the supply of electric vehicle charging stations has not kept pace with the rapid expansion of electric vehicle adoption. This incongruity raises ongoing concerns for drivers regarding Distance To Empty (DTE) or remaining range. However, accurate DTE prediction faces challenges due to various factors. Therefore, predicting effective remaining range is attracting researchers’ attention. However, most algorithms are based on long-term historical driving data, which presents limitations in a shared vehicle scenario with frequently changing drivers. To address these challenges, this paper introduces a novel algorithm employing machine learning to classify driving styles and predict remaining range. This approach can integrate expected future road information and current driving conditions, offering a solution to the uncertainties associated with traditional methods.