Predictive and transactive controls for EVE park net-zero community with AI/ML models

A net-zero community is a development that balances the amount of energy consumed with the amount of energy generated on-site, resulting in a net-zero energy consumption. This typically involves integrating renewable energy sources, such as solar panels and/or wind turbines, along with energy-efficient building designs and technologies. The Government of Canada aims at net-zero emission by 2050. In alliance with this goal, EVE Park, the first community of its kind; a net-zero energy community is being developed in west London, Ontario.

Currently the functional part of the community’s energy demand is coming from solar PV. This research work focuses on developing the AI/ML models for hourly, weekly and monthly forecasts based on the community generation and consumption data for predictive and transactive controls for the whole community. Deep learning models like RNN, LSTM and GRU are developed for this purpose. The goal is to minimize GHG emission, curb peak demand during peak hours, and reduce electricity costs. In future investigation wind energy and battery storage will be added with solar PV for further optimization.