节点文献

湖北省人口中长期变化趋势灰色组合模型预测研究

A Study on the Middle-long Period Population Trend of Hubei Province Based on the Gray Combined Forecast Method

【作者】 陈昉源

【导师】 彭定祥; 张世松;

【作者基本信息】 华中农业大学 , 农业推广, 2006, 硕士

【摘要】 人口问题是社会经济发展最根本的问题,也是所有社会经济问题决策之基础,准确地把握和预测人口变动特征与趋势,做出与人口发展相适应的判断,具有重要的现实意义。建国以来,湖北省人口增长经历了四个阶段,人口总量由1949年的2580多万人增加到2004年的6016多万人,期间有两个生育高峰期,自然增长率均达到1.92%,在第二个生育高峰期的年均增长人口高达到67万多人,总人口突破6000万人,进入全国人口大省。随着国家生育政策的调整,生活的改善,人均预期寿命的提高,湖北省人口增长发生了明显变化,年均增长率逐步趋于稳定,存在减缓态势,在这一趋势之下,湖北省未来10-15年人口变动趋势如何,缺乏数理预测分析。目前预测的方法虽然很多,但大多采用单一模型,比较简便,问题在于不同单一模型之间所得结果差异较大,可能造成一些有用信息的丢失,影响结果的准确性。本文采用定量研究方法,根据湖北省人口变动的历史数据,寻找最佳拟合曲线和方程(模型),然后通过趋势外推开展组合预测分析。组合预测分析是近年来发展较快的一种综合预测方法,它是设法把不同预测模型、方法组合起来,综合利用各种预测方法提供的信息开展预测。目前组合模型理论和应用趋于成熟。组合模型的关键在于单一模型及权重的确定。在论文中选择GM(1,1)、4次多项式、对数、乘幂等4种单一模型,分别对湖北省人口变动趋势进行预测,在此基础上,按标准差法确定单一模型在组合预测模型中的权重,建立组合模型,进行预测。分析表明,4种单一模型均能较好的拟合时间系列截面数据,对于长期趋势预测上4次多项式明显偏大,组合模型在平均相对误差、精度、残差标准差等方面表现出较好的统计结果。经过分析,在今后相当长时间内,湖北省人口在一定时间内将继续维持低自然增长率,人口总量到2010、2015、2020年将达到6185.85-6271.55、6647.07-6752.73、7715.38-7855.94万人。

【Abstract】 The population issue is essential of the society and economy development, and itis also the base of all society and economy decision-making. It has realisticsignificance to forecast the population trend and change character correctly and makeout correct judge that matches the population change trend accordingly.Since 1949, the population growth of Hubei province has been undergoing fourstages, during which time the total population has increased from more than 25.8million in 1949 to more than 60.16 million in 2004. From 1949 to 2004, we havewitnessed two high birthrate stages, in each the population accrual rate reached1.92%. During the second high birthrate stage, the total population increased by morethan 670 thousand annually, which makes the total population of Hubei pierced into60 million and hence became a big population province of our country.Along with the population policy’s adjustment of our country and theimprovement of people’s living standard, the expected life-span of people hasbecome longer than ever before and the population growth character has changed a lot,the annual growth rate will level off or even slow down. Under this situation, it isnecessary to drop out the population change trend of Hubei province in the coming10-15 years.In this dissertation, the author takes the way of ration research and adopts thehistorical population data to drop out the trend curve and the model of the populationchange of Hubei province, and the combined forecast have been done by extrapolatingthe change trend.The combined forecasting analysis is a kind of synthetic forecasting method, inwhich all different kinds of forecasting models and methods are assembled and usedto do the information forecasting work. The theory on the combined forecastingmodel and its application is popular at the present time. The key of handling thecombined forecasting model is that the single model should be good and the weight ofthe single model should be endowed correctly.In this dissertation four single models, GM(1,1)、4 times multinomial、logarithmand power model, are chosen to forecast the population change trend of Hubeirespectively. On this base, the weight of each single model is endowed by way ofstandard distance. From the analysis, we can see that all four single models can draftthe time-series data well, but the 4 times multinomial does not suit well when it isused for the long- term trend forecasting. On the contrary, the combined forecastmodel shows good statistic outcome in the average relative error、precision and thestandard error of the remnant.From the forecast analysis we draw a conclusion that the population of Hubeiprovince will maintain a low accrual rate for a rather long time in the coming yearsand the total population of Hubei province may reach the level of 61.8585-62.7155、66.4707-67.5273、77.1538-78.5594 million by the year of 2010、2015、2020respectively.

  • 【分类号】C924.2
  • 【被引频次】12
  • 【下载频次】891
节点文献中: 

本文链接的文献网络图示:

本文的引文网络