Application of Generalized Regression Neural Network in Short-term Load Forecasting

Abstract

Forecasting short-term load of electric power system is the premise of dispatching power system. The higher accuracy of load forecasting can be enhanced the utilization rate of power generation equipment and the effectiveness of economic regulation. The paper refers to generalized regression neural network (GRNN) approach to forecast the short-term load. Test samples are power load data of July 20thand weather characteristic value of July 21st, the output of GRNN is the forecast value of power system load in the next day of July 22nd.Through comparison with the prediction results of BP neural network, the experiment results show that the GRNN prediction method can be reasonably considered the short-term daily load and effectively improves the prediction accuracy.