Based On Data Mining Of Short-term Load Forecasting Researeh And Online Program Realization

Power system short-term load forecasting not only by the impact of the load sequence, but also by natural factors, date type factors, political factors, and emergency and other non-load factors. Randomness, complexity and non-linear characteristics of a variety of non-load factor, leading to large and complex power system data, which increases the difficulty of load forecasting is very important, so choose a good data mathematical methods. In this paper, the method of data preprocessing, data clean-up, conversion and cluster analysis to ensure the reliability and validity of the data, load forecasting ready. Full account of the impact of the load factor and non-load factor prediction, the use of C4.5decision tree algorithm in the Weka software on load forecasting model instance through the city in Jiangxi analysis to verify the accuracy of the requirements of the load forecasting.In this paper, from the point of view of the object-oriented analysis and design, the short-term load forecasting system requirements and functional, a relatively short-term power load forecasting system developed in Visual C++6.0platform.