Optimization of regional net load prediction based on integrated algorithm

With the increase of national electricity consumption year by year and the continuous development of electrical equipment, a large number of new power loads are connected to the grid, which puts forward higher requirements for the carrying capacity of the power system. Distributed generation has penetrated into every part of the power system. Compared with traditional load changes, today's user-side load becomes more difficult to predict and involves more influencing factors. When forecasting the net load power of a certain region, the influence of distributed generation on the net load of the region should be considered. In this paper, the Adaboost ensemble learning method is adopted to introduce the influence factors of distributed generation into the prediction of regional netload, and Adaboost processes each LSTM neural network learner by assigning weights, etc., to form a strong learner for regional net load power prediction. There are more factors involved, and the predictions are more accurate.