Intelligence Technology Applied Research, Load Forecasting Of Electric Power Enterprises

The load forecasting is an essential component in the economic and reliable operation and planning of an electric utility power system. The accuracy of the load forecast has a significant impact on the electric utility抯 operations and the development of national economy. Substantial efforts have been made to develop effective approaches for solving Short- term Load Forecasting (STLF) problem in the last few decades. For example Time-series model, it is assumed that the load profile is stationary and linear. However weather information, festive season and other factors all contributed to the load behavior, and make it of unsteady and nonlinear. So the results of forecasting often can抰 meet the need of utilities. This paper makes a study of many application problems about intelligence technology and load forecasting. Combined with an actual subject, an Artificial Neural Network (ANN) model that is sensitive to weather is proposed. To reduce the error of forecasting, a Self-adapting Fuzzy On-line Control is applied. In addition, an Expert System Knowledge-based (ES) is built in order to correct the effect from rain or temperature factors. The results of simulation and actual application demonstrated its effectiveness. The application including many intelligent methods not only provides a more accurate load forecast but also will bring about obvious economic and social benefits for the utilities.