Volume 11: Sustainable Energy Solutions for Changing the World: Part III

Energy demand estimation method for a charging station at workplace L. Bartolucci, S. Cordiner, V. Mulone, M. Santarelli, F. Ortenzi, and M. Pasquali

Abstract

Energy demand increase due to large deployment of electric vehicles combined with volatile decentralized renewable energy production is bringing up new challenges in the transmission network. Power quality issues might be avoided taking advantage from the flexibility offered by the charging process to match the local renewable energy production. However, the potential benefits from a controlled electric vehicle charging process could be optimally exploited only if electric vehicles energy demand is reliably evaluated.
This study proposes a detailed methodology to evaluate the load of a working place charging station, in order to further optimally design a second life battery storage system for ancillary services provision. In details, the electric vehicles energy demand has been estimated using a multiple linear regression model that links the vehicles battery energy consumption with microscopic driving parameters (such as speed and acceleration). In particular, the model inputs are typical driving cycles performed by the employees to reach the working place. These representative speed profiles have been reconstructed with a Markov chain-based method using real-world collected data.
The proposed approach allows to predict the battery energy consumption with a Mean Absolute Error less than 18% and with a correlation coefficient R2 of 99%.

Keywords Electric Vehicles, Energy Consumption Prediction, Markov chain theory, Multiple Linear Regression