Non-invasive Industrial Load Identification Based on KM-LSTM-AE Method
Non-invasive Load Monitoring(NILM)is widely used in non-invasive load decomposition.In or-der to solve the limitation of NILM static modeling and low identification accuracy,this paper proposes a KM-LSTM-AE method combining LSTM autoencoder and K-means clustering data cleaning,which is suitable for NILM in the industrial field.The real energy consumption data of a factory was selected as the test data set of KM-LSTM-AE.The experimental results show that KM-LSTM-AE method has high accu-racy and short time consuming.
万方数据
维普
