摘要
基于目前我国人口数量预测方法繁多,优缺点不一的现状,以盘锦市1978～2008年人口数据为基础,分别采用曲线回归、GM(1,1)与等维递补灰色模型、BP神经网络与RBF神经网络、马尔萨斯模型与费尔哈斯模型、宋健模型与Leslie矩阵方法分别对盘锦市2030、2050、2070、2090年人口进行预测,并根据预测结果设置了盘锦市人口预测方法的多目标定量优选体系,并采用Matla,b2009b软件运用3标度层次分析法方法进行了预测方法的优选.优选结果显示在盘锦市人口预测中,径向基网络(RBF)为最优方案.
Methods of population prediction are numerous and specialized. The population of Panjin in 2030 , 2050, 2070 and 2090 was predicted with the method of curve regression, gray model and improvement gray model, Malthus model and Verhulst Logistic model, Song Jian model and Leslie matrix according to the population data in Panjin from 1978 to 2008. Multi-criteria decision system for quantitative optimization of prediction method was set based on the prediction result. 3-grade AHP（3-grade Analytic Hierarchy Process） was adopt for the optimization of prediction method with Matlab2009b software. The result shows that the Radial Basis Function network （RBF） was the optimal method for forecasting the population in Panjin.
出处
《数学的实践与认识》
CSCD
北大核心
2012年第16期158-167,共10页
Mathematics in Practice and Theory
基金
国家水体污染控制与治理科技重大专项(2009ZX07526-006)
中央高校基本科研业务费专项资金资助
