Load demand forecasting of residential buildings using a deep learning model

In smart grid and smart building environment, it is important to implement accurate load demand forecasting of residential buildings. This plays an important role in supporting the reliability of the power system, improving integration of the distributed renewable energy resources, and developing effective demand response strategies. In this study, we proposed a deep learning model to forecast the load demand of residential buildings with a one-hour resolution, while considering its complexity and variability. The proposed model has a good learning ability that can accommodate time dependencies to achieve high forecasting accuracy with limited input variables. Hourly-measured residential load data in Austin, Texas, USA were used to demonstrate the effectiveness of the proposed model, and the forecasting error was quantitatively evaluated using several metrics. The results showed that the proposed model forecasts the aggregated and disaggregated load demand of residential buildings with higher accuracy compared to conventional methods. Furthermore, the proposed deep learning model is also an effective method for filling missing data through learning from history data.