Long Short-Term Memory (LSTM) Neural Networks Applied to Energy Disaggregation

Energy disaggregation is an estimation of appliance energy usage from a single meter without the needs of sub-metering. In this paper, three models of the neural networks, Decision Tree (DT), Deep Neural Networks (DNN), and Long Short-Term Memory (LSTM) were compared and implemented for energy disaggregation. The Reference Energy Disaggregation data set (REDD) dataset was used for training and testing the effectiveness of the models. It was found that the LSTM performed better performances in MAE, RMSE, and R-Square than that of the DT and the DNN networks by at least 37.38% and 86.26% respectively.