SHORT TERM ELECTRICAL LOAD PREDICATION METHOD

摘要

A short term electrical load predication method, comprising establishing a deep belief network predication model with a four-layer network structure; training the deep belief network predication model to obtain a parameter value of each layer of the deep belief network predication model, setting an activation function of the deep belief network predication model, and gaining, by means of training and learning, a mapping relation between the input and output of the deep belief network predication model; and predicating the actual electricity power of the detected electrical load to obtain an electricity power predication result of the detected electrical load. The deep belief network predication network is introduced into electrical load power predication, and by means of the deep structure of network, the inner relationship between the input and output can be learned, which achieves a predication of the load electrical power in a period of time in future, and the predication accuracy and speed are high.