Predicting Building Energy Consumption Considering Climate Change Using 6D BIM and Machine Learning

This chapter presents a groundbreaking framework that leverages 6D Building Information Modeling (BIM) and machine learning to predict building energy consumption, taking into account the impacts of climate change. The study highlights the limitations of traditional energy estimation methods and introduces a dynamic process that integrates graphical and non-graphical data to assess energy consumption accurately. The framework comprises four key stages: producing a dataset using a 6D BIM tool, training machine learning models to predict current energy consumption, generating future climate data under Representative Concentration Pathway (RCP) scenarios, and developing a regression model to forecast energy consumption considering climate change impacts. The study employs TerMus PLUS for energy simulation and Meteonorm for generating future weather data. Four powerful regression algorithms—support vector regression (SVR), Adaptive Boosting (AdaBoost), Extreme Gradient Boosting (XGBoost), and Artificial Neural Network (ANN)—are compared, with ANN showing slightly superior performance. The framework is demonstrated through a case study of a private house in Hanoi, Vietnam, showcasing its practical application and effectiveness. The study concludes by discussing the implications and future directions for enhancing real-time data integration and exploring advanced time series models to improve accuracy in predicting building energy requirements up to the year 2100.