DATA-DRIVEN MODELLING OF THE ENERGY CONSUMPTION OF AIRPORT CITIES

Abstract

The airport city of Vienna International Airport consists of more than 125 energy-consuming objects – mostly buildings but also infrastructure. In order to ensure their energy-efficient operation, digital models are needed that predict the energy consumption that the objects should have. Multiple linear regression can provide such models if historical data is available to train regression models with. In our approach, we explore how regression models must be set up for the case of airport objects. Further, we investigate how the prediction performance of multiple linear regression can be improved by combining it with other methods.