四类预测人口方法的对比及Logistic人口生长模型的改进

Comparison of Four Kinds of Population Prediction Methods and Improvement of Logistic Population Growth Model

分析当前中国的人口形势,基于现阶段人口政策,利用1970—2014年的中国总人口数据,建立了马尔萨斯人口模型、多项式拟合模型、ARIMA模型和Logistic人口生长模型对我国未来的总人口进行预测,并将这4种预测方法进行对比,得出Logistic人口生长模型的拟合优度和预测效果更优的结论,而后又针对Logistic模型存在的缺陷分别从方程和参数的角度提出了2种改进办法,使模型的精确度进一步提高.最后预测2030年我国的总人口数将在14.2亿~14.6亿波动.

Based on the current population policy and the data of China’s total population from 1970 to 2014,this paper analyzes the current population situation in China,establishes Malthus population model,polynomial fitting model,ARIMA model and Logistic population growth model to predict the future total population of our country,and compares these four prediction methods.The conclusion that Logistic population growth model is better in goodness of fit and prediction effect is obtained.Then two improved methods are proposed from the perspectives of equations and parameters to further improve the accuracy of the model according to the defects of Logistic model.Finally,it is concluded that the total population of China will fluctuate between 1.42 billion and 1.46 billion by 2030.