Lighting energy load prediction framework using agent-based simulation and artificial neural network models

Abstract: Lighting is responsible for 17% of the total electricity consumption in commercial buildings in the United States. Investigating and better understanding the lighting energy load provides the potential for more energy-saving in commercial buildings. This study proposes a framework to predict the lighting schedule and load in office buildings by integrating an agent-based model into an artificial neural network model. A small office building is used as a case study. The results illustrated that the accuracy of the prediction model could be as high as 92.8%.