Non-intrusive Load Identification Algorithm Based on Convolution Neural Network

Non-intrusive load monitoring (NILM) has broad application prospects because of its low implementation cost and low interference to users. Load identification method is one of the major technical difficulties of non-intrusive composite monitoring. In this paper, a non-intrusive load identification algorithm based on convolution neural network in non-intrusive load monitoring mode is studied. Firstly, the local mean decomposition algorithm is used for load separation from the collected mixed signals, and the independent load features are extracted via the intelligent learning method. A convolutional neural network model capable of processing two-dimensional image data is established. The convolutional neural network model is trained as a load identification model by transforming a large number current data of typical household appliances into pictures. And with this model, load characteristics are extracted to achieve the purpose of identification. The simulation results show that the load identification based on the convolution neural network has high accuracy, fast identification speed and good generalization ability. © 2020, Power System Technology Press. All right reserved.