This article does not consider the impact of short term load of various factors, only consider the relationship between the temperature and the load. We selected three models which are BP neural network model, improved model of BP neural network, the gray model to construct predict model of load. We forecasted the electricity load of Liaoning Province on December, and certificated the accuracy of the model that we used. Based on it, we find that BP neural network forecasting techniques is better than the others. So that we can use the models to forecast the short electricity load.
|