Simulating the impact of building occupant peer networks on inter-building energy consumption

We developed an integrated inter-building physical and human network model to predict the energy conservation for an assumed urban residential block. We utilized an Artificial Neural Network to predict hourly energy consumption in both the first physical and second human stage. In the first stage, simulated data were exported from EnergyPlus, and the optimal scenario was found to consume 12.28% less energy than the base scenario. In the second stage, the human network closeness index was obtained from a residential experiment to represent occupants' network connections. We found that energy consumption can be further reduced up to 51.75%. Finally, hour-by-hour energy consumption prediction under various levels of occupant networks was examined, and we found the block exhibits a potential of conserving 57.68% of the original energy consumption. An integrated understanding of physical and human network models on inter-building level energy consumption will enable us to better achieve energy efficiency objectives.