Non-intrusive Load Disaggregation Based on Multiple Optimization of Appliance Features and CNN-NLSTM Model

Non-intrusive load disaggregation technology can effectively mine customer-side appliance information, which is the basis for the grid to carry out interactive customer load response. In this paper, a non-intrusive load disaggregation model with multiple optimization selection of appliance characteristics is proposed. First, an adaptive sliding data window is designed for appliance operation characteristics to obtain a more complete power segment and adjust the network input and output dimensions. Second, extracting and deepening appliance features by fusing shallow convolutional neural networks (CNN) with two-layer nested long and short-term memory networks (NLSTM). Finally, experimental analysis on the REDD dataset shows that the multiple selection, deepening and reuse of appliance features significantly improve the accuracy of load decomposition while reducing training time.