不同预测模型下西藏自治区人口发展状况及对策研究

Different Methods and Population Development Forecast:A Case of Status of the Tibet Autonomous Region

针对西藏自治区人口数据的有限性、数据序列的不平滑性等特点,分别采用一元回归、马尔萨斯模型、lo-gistic模型、GM(1,1)模型等4种方法,利用西藏自治区1980—2009年统计年鉴数据进行人口预测,综合考虑各种方案预测值,确定西藏自治区2010-2030年的人口数,在此基础上探讨基于人口增长的西藏经济发展模式与对策。结果表明,几种预测模型中平均相对误差率最小的是一元线性预测模型Ⅲ,仅0.2611%,马尔萨斯预测模型Ⅲ的平均相对误差率最大,也只有2.0767%;从预测结果看,Logistic模型的预测结果最为保守,仅为339.61万人,马尔萨斯模型预测最为乐观,高达380.10万人,未来相关研究中,可应用Logistic模型预测值作为下限,马尔萨斯模型预测值为上限;为综合应用四种模型的优势,研究中综合平均四个不同模型的预测结果,则西藏2010年人口将达到294.07万人,2015年达到311.34万人,2020年328.81万人,2025年346.54万人,2030年364.57万人,未来20年总人口增长趋势逐渐加快,区域发展中有一定人口压力。

Using Tibetan Statistical Yearbook from 1978 to 2009 and four methods （linear regression,the Malthusian model,logistic model,GM （1,1） models） ,this paper tends to forecast the population of Tibet. The results show that the linear prediction model III has the minimum average relative error rate among the four prediction models, which is only 0.2611% ,and Malthusian prediction model m has the maximum average relative error rate; the Logistic model forecast result is most conservative, which is 3.40 million and the Malthus model forecast is most optimistic which ups to 3.80 million. In the future, the Logistic model＇s predicted value can be applied to as the lower limit of population projections, the Malthus model＇s predicted value can be taken as the ceiling value. On the mean value of the four models＇ predicted values from 2010 to 2030, Tibetan population in 2010 will reach to 2.94.million, 3.11million in 2015 , 3.29 million in 2020, 3.47million in 2025, and 3.65 million in 2030. With the gradually speeding up of the total population growth tendency in the future 20 years,there will be a certain population pressure in the future regional development. If we cannot find a better approach to coordinate the industrial development with resources and environment ,the pressure of regional population growth on the environment will further increase too.