부산대학교 도서관

Accurate and timely electricity consumption prediction results can not only provide calculation basis for power project investment, but also provide data support for optimal operation decision of power grid. Power demand shows different trends of change and development at different time stages, with complex changes, random occurrence and other characteristics. In order to predict the change of electricity consumption at different times, this paper uses VaR (Value-at-Risk) method to evaluate the electricity consumption at different times. VaR method is an important risk management method, which provides people with a comprehensive measurement of market risk and has been widely used. This paper first gives the relevant introduction of VaR; Then the Historical Simulation (HS) method and Extreme Value Theory (EVT) method are introduced; The VaR value of electricity consumption is calculated by these two methods, and the results are visualized and compared, and finally the effect of the two methods is evaluated.