Load prediction method considering distributed energy access of distribution network

The invention relates to a load prediction method considering distributed energy access of a distribution network, and the method comprises the steps: 1, constructing a power grid load data set, taking a part of data in the data set as subsequent training data, taking the other part of data as test data, carrying out the preprocessing of the training data in the data set, removing abnormal data, and carrying out the prediction of the power grid load data set; carrying out normalization processing on the data set; 2, adding the data set subjected to normalization processing in the step 1 into the constructed mathematical model of the long-short-term memory neural network for training; 3, comparing the predicted data value obtained by training in the step 2 with the test data in the step 1 to calculate an error, if the predicted data value obtained by training meets an error requirement, ending the training process, and if the predicted data value does not meet the error requirement, entering the next step; and