Study On Web Service-Based Short-term Load Forecast System Of Electric Power

The short-term load forecast is a key task to electric power systems' daily operation. Precise forecast plays very important role to the economy and security of power systems.It has been a long time from beginning to research the forecast method by computer to now. Currently there are many forecast model that have been applied into real and made good economic utility. Among them, the Radial Basis Function (RBF) neural network is a better model. It derived from the Artificial Neural Network and Distance Weight Regression. By comparison with typical Back Propagation model of the Artificial Neural Network, RBF have advantage of clear structure,fast convergence speed and good forecast precision. So it has became the focus of researching step by step in recent years.After introduce the basis model of RBF, there is an optimizing algorithm of it. This algorithm applies impulse and changeable training pace of Artificial Neural Network to RBF's training process. It can improve the speed and precision of convergence effectively.The traditional software system of the short-term load forecast almost is based on Client/Server model. With the fast development of Internet, the C/S model's shortcoming on Expanding ability and Flexibility become more obvious. This model is not fit in with current application's need, thus will be replaced by B/S model Progressively.Today the most popular developing platform of application based on B/S model is .Net platform pressed by Microsoft. By reason of powerful support to Web service technology, the .Net platform is the best tool to construct the short-term load forecast system. Especially to electric power corporation, it ought to depend on Web service technology to settle the problem that information's inconsistency and redundancy so as to enhance the level of data sharing and improve the sub-system's module quality. Above the .Net platform, it is very convenient to design and realize the short-term load forecast system that communicates with other sub-system by Web service. In the system, there are many calculating model for example RBF supplying for the visitor, so they can choose the model according to their need.