【摘要】 为提高预测老龄人口的精度，以残差修正GM(1,1)模型、灰色-BP神经网络模型和Logistic预测模型作为单项模型，建立了一种基于相关系数的诱导有序加权平均(IOWA)算子组合预测模型.利用该模型对2000—2020年我国东部、中部、西部和东北部地区及全国老龄(≥65岁)人口数量进行预测显示，该组合预测模型的预测效果显著优于上述3个单项预测模型，表明该模型能够有效地提高老龄人口的预测精度.利用该组合预测模型对未来10年我国东部、中部、西部和东北部地区及全国老龄人口数进行预测显示：我国老龄人口数总体呈现逐步增长态势，同时区域间的老龄化进程差异也不断增大，其中东部地区老龄人口增长得相对较快，中部地区则增长得相对较慢.该预测结果可为我国老龄人口政策的制定提供参考.更多还原

【Abstract】 In order to improve the predicting accuracy of the aging population, a combinational forecasting model which is based on induced ordered weighted averaging(IOWA) operators and correlation coefficient is established by using the residual modified GM(1,1) model, the gray-BP neural network model and the Logistic forecasting model as the single model.Using this model forecast the aging population(≥65 years old) of the eastern, central, western and northeastern regions and the whole country for 2000-2020 in China, the forecasting result shows that the predicting effect of the combinational forecasting model is significantly better than that of the above three single forecasting models, indicating the model can effectively improve the predicting accuracy of the aging population.Finally, the combinational forecasting model is used to predict the number of ageing population in the eastern, central, western and northeastern parts of China and the whole country in the next 10 years: the number of ageing population in China generally shows a gradual growth trend, and the differences in the aging process between regions are also increasing, of which the ageing population in the east grows relatively fast and the growth in the central region is relatively slow.The forecasting result can provide a reference for the formulation of Chinese aging population policy.更多还原