BP Neural Networks Algorithm Applied in Electric Load Prediction

Abstract

After analyzing the meaning and ways of electric system load prediction, this essay illustrates the general theory of the BP neural networks and study the load prediction methods based on BP neural networks. It analyzes the neural networks sample value methods, performs encumbrance for historical data, then, it preprocesses the load data. The accuracy of the load prediction results is closely related to the preprocessing of the load data. The data processed correctly will produce distinct effect on the accuracy of the prediction results. Contrarily, inputting the load data unprocessed into the neural networks will result in prediction failure. This essay designs the structure of the algorithm of BP neural networks. At last, it analyzes the data of the case study of electric load prediction.