Based on the data of the social electricity consumption from 2000 to 2014 of Anhui Province,this paper, using multiple linear regression,quadratic exponential smoothing and polynomial fitting,gives individual predictive values.Introducing GIOWA operator (generalized,induced,ordered,weighted averaging),the paper selects IOWA , IOWHA,IOWGA operators and establishes the error square and a minimum combination forecasting model.Through the evaluation model, three kinds of combination prediction models‘errors are all smaller than that of the single forecasting,which verifies the validity of the model.In the support of the established model,the total electricity con-sumption in Anhui in five years ahead can be predicated,providing a decision-making basis for power development planning and power supply-side management.%选取2000-2014年安徽省全社会用电量数据，采用多元线性回归、二次指数平滑以及多项式拟合3种预测方法，给出了单项预测值。引入广义诱导有序加权平均(GIOWA )算子，选取其3种特殊形式IOWA、IOWHA、IOWGA的算子，建立误差平方和最小组合预测模型，并通过模型评价发现，这3种组合预测模型的误差均小于单项预测值，验证了模型的有效性。利用所建模型预测了未来五年的安徽省全社会用电总量，为电力规划和需求侧管理提供决策基础。
展开▼
