Self-adaptive power load identification method and device

The invention relates to a self-adaptive power load identification method and device, and belongs to the field of power load monitoring and identification. Calculating the minimum power-on time, the minimum power-off time and a power-on threshold value; adding the obtained three parameters into a model training process, and extracting effective load activation data from the preprocessed data; based on the power load data, training the power load identification model by using a long-short-term memory artificial neural network; the method comprises the following steps of: dividing a test set according to a proportion, and obtaining an optimal'window 'capable of controlling test time within a time threshold value by using a process processing technology in a'window' searching stage; and defining the size of a sliding window in a real-time reasoning program by using the optimal'window ', and decomposing and identifying the power load data. The method is not limited by initial model training data any more, the rec