Mesoscopic model framework for estimating electric vehicles' energy consumption

Promoting the usage of electric vehicles (EVs) is viewed as an effective way to develop sustainable transportation. Accurate energy consumption estimation is the key to plan charging infrastructure and provide intelligent service for EV customers, which can further accelerate the popularity of EVs. In this paper, to make a compromise between the model accuracy and data intensity, a mesoscopic model framework for estimating EVs' energy consumption is established based on real driving condition data. First, the energy consumption characteristics of EVs are analyzed from both statistical and physical standpoints to reveal the energy consumption mechanism of EVs. Then, with the explicit consideration of the regenerative braking under decelerating mode, which is a distinctive feature of EVs, the required regenerative braking power is given special treatment and is incorporated into the modelling framework. Finally, an energy consumption factor model based on the average travel speed is established to associate the finer scale energy consumption on a link-by-link level. The proposed model is validated against with field measurements, and the results indicate the model has a good predictive ability. This study provides a more complete manner for considering unique energy consumption characteristics in EVs' energy consumption estimation under real-world driving conditions.