Karamay Area Of ​​power System Load Forecasting

Electric Load forecasting is very important to programming of electric power system, investment and production management.This thesis represents developing of the oil city, prospects the changing economy of karamay, describing trends and dynamic state of region economy and electric power system, collecting serious economic data of karamay, which includes GDP, oil production, electricity requirement, load and etc. It briefly introduced classification and characteristics of electric load , manner and process of electric load forecasting ,combine with statistical theory and methods. It attempts electric elastic coefficient methods, time series analysis and regression data models in formulating historical data. In process of comparing, less subjectivity and more objectivity is concerned, try to find the way suit to application of scientific forecasting, make actual use of production management and programming of electricity and investment.In process of study, because the caliber of original GDP data is hard to unify , electric elastic coefficient are more infulenced by subjective factor. Other way , Model's precision of electric demands that rely on oil production is low. It only suit to auxiliary local electric demands foresting. Three Exponential Smoothing model's precision is high, but the result of this model are far away from actual. In beginning, the precision of Trend projection model are low, but with regulation of time variable, the best result is available.It is more important that testing and regulating over and over than pursuing accuracy and visual effect in the beginning. In fact, it is necessary that combine qualitative analysis with quantitative analysis to make the model more actual and more theoretical.