基于曲线拟合的中国城镇人口数预测研究

A Study on the Prediction of Urban Population in China Based on Curve Fitting

分析了中国城镇人口的发展状况,选取1974年-2021年的城镇人口数据,利用MATLAB拟合工具箱建立中国城镇人口数与时间序列的多项式函数模型和指数函数模型。虽然二项指数函数的相关系数、方差和标准差均优于其他模型,但通过对未来10年数据的预测,发现该模型呈下降趋势,并很快降为零,与现实不符。最终选取三次多项式函数模型进行拟合,该模型相关系数高达0.9986,方差与标准差相对较小。拟合结果显示中国城镇人口数会持续增加,到2035年增长为131090万人。

Analyzed the development status of urban population in China,selected urban population data from 1974 to 2021.Establishing Polynomial and Exponential Function Models of urban population and Time Series in China Using MATLAB Fitting Toolbox.Although the correlation coefficient、variance and standard deviation of the binomial exponential function are better than other models.Through predicting the data for the next 10 years,it is found that the model shows a downward trend and quickly drops to zero,which is inconsistent with reality.This article ultimately selects a cubic polynomial function model for fitting,with a correlation coefficient of up to 0.9986 and a relatively small variance and standard deviation.The fitting results show that the urban population in China will continue to increase,reaching 1310.9 million by 2035.