Similitude approach for buildings energy forecasts

Buildings are assets characterized by environments and uses that change over time, variable occupancies, and long life cycles. In addition to this, they have high operational costs, mostly due to their energy demands, accounting for 30% to 40% of global greenhouse gas emissions. All of this makes them as complex as challenging to manage in an efficient way. Consequently, increasing efforts have been made to forecast their energy needs, with the scope of optimize their economic and environmental impact. About this, the available literature focused mainly on short-term modeling through the implementation of sets of physics-based equations (i.e., whitebox), functional relationships between input and output variables (i.e., black-box), or a combination of both (i.e., grey-box). Far from reaching a state of art condition, more research is necessary on long-term forecast models, especially with the aim of reducing the energy needs than optimize the energy production. Within this context, this thesis presents an original automatic and integrated similitude approach for forecasting the energy needs of buildings from short to long-term time horizons. This is accomplished by scaling an unknown facility from a similar facility that is already known and by executing a black-box approach based on machine learning algorithms. The proposed method is implemented in real case studies in Italy to predict the energy demands (i.e., heating, cooling, and electricity) of Sant’Anna Hospital in Ferrara through the historical data of Ca’ Foncello Hospital in Treviso. The results show an adjusted coefficient of correlation above 0.7 and an average error below 10% for all the energy demands, demonstrating a feasible forecast performance with a low training set-to-test set ratio.