Multi-element load prediction method for integrated energy system based on machine learning

The invention discloses a multi-element load prediction method for an integrated energy system based on machine learning, and the method comprises the steps: carrying out the processing of the possible incompleteness or abnormality of historical load data, completing the elimination of abnormal data through employing the Pauta criterion, completing the data completion through employing a weighted sequence filling method, and carrying out the prediction of the abnormal data. Determining related influence factors in a load prediction process by using a Pearson's correlation coefficient, performing normalization processing on the data after the influence factors are determined, screening the data based on a similar day theory, dividing the processed data into a training set and a test set, and constructing an improved intelligent water drop algorithm to optimize a deep belief network model; and transmitting the training set to a deep belief network to obtain a load prediction model, and transmitting the test set