基于MATLAB的人口预测研究

发布时间：2018-06-10 18:41

本文选题：人口预测 + MATLAB　；参考：《华东师范大学》2010年硕士论文

【摘要】：本文是采用MATLAB仿真计算平台对人口预测的应用问题所进行的一些研究,其研究工作基础与研究背景源自本人承担的2010年浙江省教育厅科研课题项目《基于MATLAB仿真预测模型及犯罪数量实测研究》的前期工作、2009年浙江警官职业学院科研课题项目《基于MATLAB曲线拟合分析研究》与《基于MATLAB人口预测分析研究》以及多年来从事《数学实验》的教学工作。相关的研究成果已发表于吉林师范大学学报(自然科学版,2010年5月)及微计算机信息学报(2010年8月) 本论文的研究成果主要包括：采用MATLAB软件仿真实测出近期人口预测模型：以MATLAB仿真计算平台,以中国1990-2008年历年人口统计资料为数据环境,仿真实测出了适于近期人口预测的简洁高效的多项式模型与傅立叶级数模型。依照预测时间的不同,从多个角度探讨了近期预测模型的准确度,进而得到了具有较高准确度的近期人口预测模型。改进了传统LOGISTIC人口预测模型的相关参数：以往传统的中远期人口预测模型多引自于荷兰学者Verhaus提出的LOGISTIC模型,其数理上的人口总量并非一个无限增长的数学过程,内含的人口阻滞作用将使人口达到饱和而最终趋于常量,但人口预测的准确度取决于如何选取模型及待定参数。为解决这个问题,本文采用MATLAB仿真计算平台进行了实测分析与研究,对LOGISTIC人口预测模型进行了泛化式改进,获得了与其它文献不同的人口环境容纳量。由本文完成的MATLAB仿真实验表明,改进后的LOGISTIC人口预测模型准确度明显提高,预测的平均相对误差由4.485%降低为0.606%,对LOGISTIC模型预测值再次修正后,预测的最大相对误差由1.30%降低为0.38%,其精度较好满足中远期人口预测的要求。依所建立的LESLIE模型编程实测而获取了人口结构的重要数据：以66岁及以上为老年人口这个新的界定标准,得出了50年间各年龄段的人口数量,并进一步修正了LESLIE模型的预测误差,提高了近期人口预测的准确度。本文编程实测所获取的人口结构数据表明,未来50年中国人口扶养比攀升加快,仅基于2001年8.11%的中国人口扶养比进行预测,中国人口扶养比在2042年将增至32.06%,可能提前显现发达国家的人口结构,将给人均资源匮乏的中国社会造成沉重的负担。
[Abstract]:In this paper, some researches on the application of population forecasting are carried out by using MATLAB simulation and calculation platform. The foundation and background of its research work come from the preliminary work of the 2010 Zhejiang Provincial Education Department Scientific Research Project "Simulation and Prediction Model based on MATLAB and the Measurement of Crime quantity", and the 2009 Zhejiang Police officers Vocational College scientific research. The subject items "the Analysis and Analysis of Curve fitting based on MATLAB" and "the Analysis of population Prediction based on MATLAB" and the Teaching work of "Mathematical experiment" for many years. The related research results have been published in the Journal of Jilin normal University (Natural Science, May 2010) and the Journal of Microcomputer Information (August 2010). Port prediction model: MATLAB simulation calculation platform, Based on the demographic data from 1990 to 2008 in China, a simple and efficient polynomial model and Fourier series model suitable for the recent population prediction have been simulated and measured. According to the different prediction time, the accuracy of the short-term prediction model is discussed from several angles. Furthermore, the recent population forecasting model with high accuracy is obtained, and the parameters of the traditional LOGISTIC population forecasting model are improved: the traditional medium-long term population forecasting model is derived from the LOGISTIC model proposed by Verhaus, a Dutch scholar. The mathematical total population is not a mathematical process of infinite growth. The population blocking effect will make the population saturation and eventually tend to be constant, but the accuracy of population prediction depends on how to select the model and the parameters to be determined. In order to solve this problem, this paper uses MATLAB simulation calculation platform to carry on the actual measurement analysis and the research, has carried on the generalized improvement to the LOGISTIC population forecast model, has obtained the population environment capacity which is different from the other literature. The result of MATLAB simulation shows that the accuracy of the improved LOGISTIC population prediction model is obviously improved, the average relative error of the prediction is reduced from 4.485% to 0.606, and the prediction value of the LOGISTIC model is revised again. The maximum relative error of forecast is reduced from 1.30% to 0.38, and its precision is good to meet the requirement of medium and long term population forecast. The important data of population structure are obtained according to the established Leslie model, which is the elderly population aged 66 and above. This new definition standard, The population size of each age group in the past 50 years is obtained and the prediction error of Leslie model is further corrected to improve the accuracy of population prediction in the near future. The population structure data obtained by programming and measuring in this paper show that the population dependency ratio of China will increase rapidly in the next 50 years, only based on the forecast of 8.11% of the population dependency ratio in 2001. China's population dependency ratio will increase to 32.06 in 2042, which may show the population structure of developed countries ahead of time, and will impose a heavy burden on the Chinese society, where per capita resources are scarce.
【学位授予单位】：华东师范大学
【学位级别】：硕士
【学位授予年份】：2010
【分类号】：C921

【引证文献】