Combination and optimization load forecasting based on credibility

Abstract

Load forecasting is based on the known history of electricity demands, and it considers political, economic, and climate factors to make forecasting for the future electricity demands. This article first analysis the trends of history power load, and then choose load forecasting methods according to the trends. Through the analysis and comparison of the load forecasting methods’ credibility, assess the adaptability between forecasting methods and power load integrated. Higher credibility load forecasting methods are got through credibility sorting and then weights for each method are solved by optimization. The sorted load forecasting methods are combined together to forecast load power. Then to make the forecasting result more accuracy, after-test residue checking method is used to test the forecasting results. Forecasting electricity consumption of local actual operation power load and getting high credibility forecasting results can ensure the construction of local power grid develop with correct direction.