Digital Twin Framework for Powertrain Energy Consumption of Fuel Cell Electric Bus

This paper proposes a digital twin framework for analyzing the traction energy consumption in fuel-cell electric buses (FCEBs). The proposed approach combines real-time sensor data and Internet of Vehicle (IoV) devices on FCEB to build a digital replica of their traction systems, including powertrain, energy storage, and energy management. Digital twins with machine learning capabilities emulate energy consumption patterns and identify potential areas to minimize Hydrogen consumption. In the specific FCEB case study and simulations, the proposed framework provides insight to improve the operational performance of FCEBs. The best-developed AI model achieved a low consumption prediction mean squared error of 0.4439%.