Low-income load profile development using occupancy and occupants' activities in residential households

Occupants' activities and dynamic presence together with characteristic variables linked with energy loads/usage within domestic dwellings are not reflected in many applications presently implemented in modeling load and energy profiles. This study makes use of an artificial neural networks system (ANN) for load and energy profile prediction. For this investigation characteristic variables such as active occupancy, occupants' interactions with residential buildings, and low-income level are considered. The performance of the proposed forecasting model in terms of low-income recipients category was evaluated based on statistical measures. Based on the outcome, the model presented a good correlation and mean square error (MSE).