Modelling community-scale renewable energy and electric vehicle management for cold-climate regions using machine learning
Rahim Zahedi, Mohammad Hasan Ghodusinejad, Alireza Aslani, Caroline Hachem-Vermette
Abstract
With increasing environmental problems of fossil fuel-based devices and systems in societies, diffusion and adoption of sustainability solutions such as renewable energy technologies and hybrid/electric vehicles have increased in the residential and commercial sectors. However, energy demand in the buildings and energy supply from renewables are complex, dynamic, and non-linear. This complexity shows itself in cold-climate regions that the supply of required energy from renewables is along with uncertainties and even sometimes stochastic. This paper assesses the energy supply/demand performance of a group of residential buildings in a community in a cold-climate region, St. Albert, Canada. First, all the buildings of the community are modeled and the required energy to respond to the demand and electric vehicles are calculated. Then, the potential of each building for electricity supply via photovoltaic panels is calculated. Finally, the energy supply/demand management of the community is assessed using a machine learning tool. The results show that the community peak heating and cooling loads are 420 kW and 121 kW, respectively and the total annual energy production from photovoltaic system is 14,203 kWh for a single house. Regarding the electric vehicle load, the photovoltaic system can provide 29.23% of the community's total load, annually. Finally by comparing the modeled pattern and the predicted pattern, an accuracy of 88.6% is obtained for the prediction.