A comparative study of vehicle powertrain efficiency: Data-driven analyzing energy consumption and environmental impact
Witsarut Achariyaviriya, Wongkot Wongsapai, Damrongsak Rinchumphu, Nakorn Tippayawong, Korrakot Yaibuathet Tippayawong, Pana Suttakul
Abstract
• Compares real-world energy consumption of BEV, HEV and ICEV. • Utilizes ML algorithms for detailed energy usage analysis. • Demonstrates BEVs’ superior environmental benefits over time. • Proposes policy recommendations to support BEV and HEV integration. • Identifies key factors affecting energy consumption using SHAP approach. This study investigates the energy consumption of electric vehicles (EVs) in real-world driving conditions, using in-vehicle sensor data and location tracking. The energy consumption of EVs is compared to that of an internal combustion engine vehicle (ICEV) under similar driving tests. Furthermore, an energy-centric life cycle assessment was conducted to assess the environmental impact of the transition to electrified transportation by reporting the carbon emissions of the EVs and ICEV. Since a large volume of driving data was acquired in the experiment, machine learning (ML) techniques are employed to interpret the obtained data and identify a key feature. ML algorithms can help develop predictive models that estimate energy consumption and explain the major factors that impact such consumption. The findings also offer valuable insights into energy consumption, emission, and the underlying factors that can inform policy discussion related to transport electrification and enhance energy efficiency. This research is significant in comprehending the potential of current EVs and their superiority over traditional vehicles. It is vital to facilitate the shift toward electric transportation, which can be driven by the results.