Short-Term Load Forecasting Of Small Load Heavy Wave Area Model Research

Short-term load forecasting (STLF) takes an important act in dispatching, distributing and programming of EMS (Energy Management System) in power system. The degree of load forecast influences the economical and social benefit directly. With the electric market developing and power demand increasing quickly, STLF is not only restricted to be a part of areas with heavy load, but also be a necessary part of areas with small load such as a county, so that to use energy more efficiently. These areas are called small load heavy wave area (SLWA) in the thesis, whose load waves heavily with some proportional impulsive load existing. Now, there are only a few of reports about this research and most of them are referred to long-term forecast home and abroad.The thesis made a primary research in STLF of SLWA. By analyzing the characteristics of SLHA it concluded that heavy wave was the most character. Therefore the thesis introduced wavelet transform modular maximum to process the history data, for one thing it can identify the abnormal data, for another can smooth noise of the data. Then the thesis built two different models for summer and other seasons by BP network. Furthermore, for BP network can't make an accurate response when the training pattern changed, the thesis proposed two methods to modify the BP model of summer and got better precision. The experimental results showed that the models had good precision and was useful for STLF with the characters above.