The Optimized GM(1,1) Load Forecasting Model Based on Data Fusion Algorithm

In order to improve the precision of load forecasting of the medium-and long-term power and to solve the problem of accuracy reduction of traditional grey GM(1,1) model caused by the contingency deviation of the original background data in the medium-and long-term load forecasting,this paper suggests a solution to combine the data fusion algorithm with the GM(1,1) model so as to form an optimized GM(1,1) model based on the data fusion algorithm.First of all for a particular year,a number of different historical background data are taken for GM(1,1) model prediction.Then data fusion algorithm can be used to make optimization analysis on multiple predicted value and the optimized forecasting results can be obtained.Finally,through the case analysis of the annual electricity load of a power system,it can be proved that the optimized GM(1,1) model based on data fusion has higher prediction accuracy and that it has a good predictive ability for the medium and long term load in the electricity system.