Energy Consumption Estimation for Routing EVs based on Driver Behavior

There has been a significant increase in the sales of electric vehicles (EVs) in the United States and abroad in the last few years. Nevertheless, the overall adoption of these vehicles is hindered by range limits of EVs in conjunction with long charging times. In this context, it is essential to determine current energy demands and to predict future demand. This paper presents approaches for predicting energy consumption of EVs and discusses their eligibility for this purpose. Four modeling approaches (i.e., dynamic systems, neural networks, statistical models, physics-based models) have primarily been used in recent literature. In order to predict an EV’s energy demand, several modeling techniques are combined to give an accurate prediction of the future energy consumption. For the battery EVs a combination of physics-based modeling and statistical modeling have shown to be an effective and efficient choice.