Multi-level non-intrusive load identification based on k-NN

Non-intrusive load monitoring (NILM), which is an important part of intelligent power system, has great research significance. In this paper, a two-layer classification method based on k-nearest neighbor(k-NN) algorithm for nonintrusive load identification and disaggregation is proposed. The active power and reactive power of the load are taken as the load characteristics identified in the first layer, and then the third and fifth harmonics of the current are taken as the load characteristics identified in the second layer. By k-NN algorithm, the load to be disaggregated is compared with the data in the training set to realize the identification and disaggregation of the load.