收藏设为首页 广告服务联系我们在线留言

干旱区地理 ›› 2013, Vol. 36 ›› Issue (5): 955-962.

• 区域发展 • 上一篇    下一篇

基于多模型的西宁市人口规模预测

张海峰,杨萍,李春花,周强,高丽文,陈琼   

  1. 青海师范大学生命与地理科学学院, 青海    西宁    810008
  • 收稿日期:2012-11-02 修回日期:2013-02-27 出版日期:2013-09-25
  • 作者简介:张海峰(1969-),男,甘肃庆阳人,教授,博士,北京大学访问学者,主要从事区域发展与规划管理及GIS应用研究.Email:haifzhang@126.com
  • 基金资助:

    教育部春晖计划资助项目(S2011008、Z2011023);青海省科技厅应用基础研究计划项目(2011-Z-741)

Population prediction of Xining City based on multi-models

ZHANG   Hai-feng,YANG  Ping,LI  Chun-hua,ZHOU  Qiang,GAO  Li-wen,CHEN  Qiong   

  1. College of Life and Geography Science of Qinghai Normal University,Xining 810008, Qinghai, China
  • Received:2012-11-02 Revised:2013-02-27 Online:2013-09-25

摘要: 人口是反映国情、国力基本情况的重要指标,是区域研究所必须考虑的重要因素之一。合理、准确地预测城市人口规模,是城市与区域规划中首先要考虑的基本问题,也是保证规划科学性与可实施性的关键性前提。以西宁市2000-2011年历年总人口为样本数据,分别构建了一元线性回归模型、马尔萨斯模型、logistic模型及GM(1,1)模型,并进行模型检验。结果表明:(1)模型均通过模型精度检验且精度较高,GM(1,1)模型拟合度最高,均误差达到0.004%,马尔萨斯模型拟合度最低,为-1.440 8%;(2)分析模型预测精度差异产生原因及适用性,表明深入、准确地分析样本数据特征,恰当选择分析方法对于控制人口预测精度尤为重要。由于西宁市2000-2011年人口样本数据在2005及2009年数据存在波动性,破坏了其与一元线性回归模型及马尔萨斯模型的拟合度,导致在4种模型中,Logistic及GM(1,1)模型预测精度较高,而GM(1,1)模预测精度最高,所以采用GM(1,1)模型进行西宁市人口预测,得到西宁市人口预测的最终结果:2012年西宁市总人口将达到225.89×104人,2015年将达到233.39×104人,2020年将达到246.37×104人。从结果看,未来9 a西宁市人口将呈现持续平稳增长的态势,但随着时间推进人口增长速度将逐渐下降。

关键词: 人口规模, 趋势外推法, 马尔萨斯模型, logistic 模型, GM(1, 1)模型, 西宁市

Abstract: The population system is a dynamical system. The population is a very important factor that affects regional economic development,the trend of a population will affect the development of the society and its economy,the accurate population prediction provides a scientific basis for regional planning and decision-making,so it has great theory meaning and realistic meaning to exactly predict development trend of population and establish rational population layout. The model selection of population prediction and scientific confirmation of prediction parameters are two most important basic steps. In this paper,according to 2000-2011 annual statistical data of the total population of Xining City,Qinghai Province,China,the unary linear regression model,Malthusian model,logistic Model and GM(1,1)Model are gained by the least square method. Theoretical and historical simulation examinations show that four models fit well,so these models are used to simulate and forecast the total population from 2012 to 2020 in Xining City. The conclusions indicate as follows:(1)All of the models(the Malthusian model,linear regression model,logistic model,and GM(1,1) model) passed the check for population prediction,and the match accuracy is high;(2) The relative errors of the prediction results of the models are calculated,and verified forecasting accuracy of the models by using a testing value. The results show that the accuracy of predicting values by the models for population prediction in Xining City is above 98%,the mean errors of the unary linear regression model,Malthusian model,logistic model and GM(1,1)model are -0.771 8%,-1.440 8%,0.120 8% and 0.004%,so GM(1,1)model is better than other models according to mean errors;(3)Because that GM(1,1)model is better than the others,so this model is used in population prediction of Xining City:it will reach 2.258 9 million people in 2012,2.333 9 million people in 2015 and 2.463 7 million people in 2020.

Key words: population scale, trend extrapolation, Malthusian Model, Logistic Model, GM(1, 1)Model, Xining City

中图分类号: 

  • C924.24