Power load real-time classification decomposition method and system based on deep learning

The invention relates to a power load real-time classification decomposition method and system based on deep learning, and belongs to the technical field of power load monitoring identification, and the method comprises the steps: calculating the minimum power-on time, the minimum power-off time and a power-on threshold value of each electric appliance; preprocessing the data; training a power load classification model; training a power load decomposition model; generating a power load identification reasoning program according to the power load classification model and the power load decomposition model, and uploading a reasoning result to a database of the cloud platform to realize information interaction with the applet; according to the invention, the edge device is utilized to classify and decompose the power load, and the accurate decomposition and monitoring of the power load are realized by establishing an applet back-end web service on the cloud server and interacting with a user through an applet; a