Executive Summary : | The building sector is a significant contributor to global energy consumption, accounting for 55 percent of energy use and 28 percent of CO2 emissions. To achieve net-zero emission goals by 2070 and 2050, India and Canada need to transition to less energy-intensive buildings, such as net-zero energy buildings (NZEBs). Building Energy Simulation (BES) tools are commonly used to predict building performance, but they are computationally intensive. To overcome this, researchers have turned to Machine Learning (ML)-based surrogate models to complement BES and overcome computational limitations. ML models can be trained and tested to mimic BES capabilities, allowing for the testing of many building design scenarios at a low computational cost. The proposed project aims to address these gaps by proposing a Simulation-based Machine Learning Optimisation (SML-Opt) framework for NZEB design and operation, focusing on retrofitting existing buildings. The project will demonstrate and validate the framework on buildings in Canada and India, providing recommendations for adjusting building regulations to support the net-zero transition. |