Analysis Of Power Load Characteristic And Study On Medium And Long-Term Load Forecasting

Analysis of power load characteristic needs to be considered with effects of many factors. Medium and long-term load forecasting model also needs to reflact the relationship between these influence factor and power load.In this study, multivariate statistical analysis reveals the inner relationship between electric load growth and several factors after analysis of main power load characteristic index. Using principal component analysis, the existing economic indicators are replaced by the original target of decreasing PCA, effective relief to the dominant factor.Electricity demand is closely related with economic development. In this paper, periodic fluctuation theory is used to analyze the correlation between electricity demand and features of economy. According to historical data between 1975 and 2006, Shanghai electric load changes divided into five cycles. We analyse each cycle fluctuations parameters and stability. Relationships between electricity demand and economic policy are revealed to provide basis for the trend of electricity demand forecasting.Traditional methods of load forecast are often based on the characteristic of historical load array itself, or use definitude relationship to simulate influence of several relative, which leads to flaws of grasping the regularity of power change. In the field of the information, MI is used to analyze the dependence level of two random variables. It contains linear correlation and non-linear correlation and can be used to describe the relationship of load and its relative elements, which provides load forecasting's basis. According to MI, this paper builds MINI for load forecasting. This model features, compares, and chooses those relative elements step by step, and then receives the relationships between each economy index and load. Using the model to calculating a real power system, satisfactory forecast result has been got.