Power demand side distributed load regulation and control method and system based on federated learning

The invention provides a power demand side distributed load regulation and control method and system based on federated learning, and the method comprises the steps: collecting the load state information of an adjustable load unit, encrypting the load state information through a load layer, obtaining the encrypted load state information, and uploading the encrypted load state information to a terminal layer; the terminal layer decrypts the encrypted load state information and then establishes a training data set; aiming at an adjustable load unit in the regulation and control area, the main station layer establishes a regulation and control model based on a regulation and control target; the terminal layer and the master station layer jointly train the regulation and control model based on federated learning until the model converges, and a convergent regulation and control model is obtained; the master station layer generates a regulation and control instruction according to the convergent regulation and con