Power Short-term Load Forecasting Based On Artificial Neural Network System

Nowadays, our country is being in a serious electric power scarce period. The electric power shortage in nation-wide scope not only has seriously affected the industrial production and brought great inconvenience to people's life, but also has caused a series of social problem. To solve the problem of electric power shortage, we must deepen on the reform of electric power market continuously. Thus, it requires the power enterprises to excavate potential of themselves under existing conditions, reasonably manage power resource, and reduce the pecuniary loss brought by the electric power shortage as best as one can.The short-term electric load forecast is a very important link in the electric power management, resulted in a great improvement in the establishment of the production plan of generators, the reasonable deploy for fire and water generate electricity and fuel, the short-term maintenance of safe analysis equipments and the spread of the energy of electrical network.. In a day that informative construction are in vigorous development, traditional artificial forecast can not already satisfy the needs of electric power industrial development more and more. However, the new forecast method of representative, which is artificial neural network, shows its huge superiority.After studying plenty of documents and analyzing various important factors of electric power load, the BP network with three-layer structure has been constructed, and the network was trained with historical electric power load data, resulted successful short-term electric power load forecast. And the precision has risen to a higher level compared with the result forecasted by traditional methods, which proves the validity of artificial nerve network in the power load forecast field of short period. After the artificial neural network model of short period power load forecast had been'established, and guided by the object-oriented analysis and design method, a simple prototype of nerve network load forecast system has been developed usingJava language, which further proved the feasibility of power load forecast system of short period based on the artificial neural network.