Smart Buildings: Comparison of Various Deep Learning Models to Forecast Energy Consumption

Smart buildings are increasing around the world as they offer a range of benefits including energy efficiency, cost savings, and improved occupant comfort. Equipped with the different innovation technologies like artificial intelligence and machine learning. Such technologies are applied to predict energy consumption in smart buildings. Deep learning, a subset of machine learning, has become more popular in recent years due to its capability of learning complex patterns from large datasets. In this paper, the application of deep learning on energy consumption prediction of smart buildings is reviewed. An overview of smart buildings and energy consumption prediction are provided. The basic principles of deep learning and its application in smart buildings are then discussed. The benefits and drawbacks of several deep learning models developed for smart building energy consumption prediction are discussed. Finally, future research directions for applications of deep learning to forecast energy consumption in smart buildings are concluded.