Population Prediction Research Based On The MATLAB

This paper is a study on the application of population prediction which adopts MATLAB simulation platform. The basis and the background are originated in the author's:preliminary work of 2010 Education Department of Zhejiang Province scientific research project "The measurement research based on MATLAB simulation and prediction model and crime quantities "; 2009 Zhejiang Police Vocational Academy scientific research project "Curve fitting analysis based on MATLAB ", "Population forecast analysis based on MATLAB" and teaching of engaging in "Mathematics Experiment" over the years. Relative research findings are published in Jilin Normal University Journal (Natural Science Edition, May 2010) and Control and Automation Journal (August 2010).The research findings mainly include in the followings:With MATLAB simulation calculation platform and data environment of population statistics information from 1990 to 2008, simulation and practical measurement which is applied to recent population forecast of polynomial model and Fourier series model have been found. According to different time predictions, the paper discusses the accuracy of recent forecast model from multi-perspectives and finally establishes recent population forecast model with a high accuracy.Improving relative parameters if classical LOGISTIC population forecast model as traditional medium and long-term population forecast models are mostly cited from Holland scholar Verhaus. In his LOGISTIC model, the total population is not an infinite growth process, but eventually tends to constant due to the population block function within it will enable population to saturation. However, the accuracy of population forecast model depends on how to select model and undetermined parameters. In order to solve the problem, the paper analyzes actual measurement and deliberates with utilization of MATLAB simulation platform, makes generalized improvements to LOGISTIC population prediction model was improved, and achieves environment capacities of population which are different from other literature. MATLAB simulation experiments show that the accuracy of improved LOGISTIC population prediction model has been obviously enhanced, with an average relative error from 4.485%to 0.606%. In addition, after re-modifying of and LOGISTIC predicted value, the maximum error is reduced from 1.30% to 0.38%, which better meets the demand of middle and long-term population prediction.Based on LESLIE model, the paper measures important data of population structure. With a new demarcation standard of aging population of 66-year-old and above, the paper finds population of different age groups in 50 years, further modifies prediction error of LESLIE model and the raise the accuracy of recent population forecast. The population structure data originated from programming and actual measurement in the paper reveal that dependency ratio of population in China will ascend sharply in the next 50 years. Only predicting from 8.11% in 2001, China dependency ratio will soar to 32.06% in 2042, which interprets population structure of developed countries will probably appear in advance. This will bring heavy burden to China's society which lacks of per capita resources.