Sag type identification method, device and system based on deep learning SDAE-B

The invention discloses a sag type identification method, device and system based on deep learning SDAE-BP, and the method comprises the steps: obtaining voltage sag waveform data of a to-be-identified sag type, and carrying out the preprocessing; inputting the preprocessed voltage sag waveform data into a pre-trained stack noise reduction auto-encoder (SDAE) to obtain voltage sag feature data extracted by the SDAE, and inputting the voltage sag feature data into a pre-trained BP neural network model; and finally, determining the sag type of the voltage sag waveform data according to the output of the BP neural network model. According to the method, the stack noise reduction auto-encoder (SDAE) in deep learning is used for extracting the voltage sag signal features, then the BP neural network is used for identifying the sag type, the problem that manual feature extraction is affected by unknown features and noise can be solved, and the accuracy of sag type identification is improved. 本发明公开一种基于深度学习SDAE-BP的暂降类