时滞投入占用产出技术研究

项目名称： 时滞投入占用产出技术研究

项目编号： No.61273208

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 自动化技术、计算机技术

项目作者： 陈锡康

作者单位： 中国科学院数学与系统科学研究院

项目金额： 72万元

中文摘要： 投入产出分析是系统科学和经济研究中常用的一种方法。其创始人Wassily Leontief曾获得1973年诺贝尔经济科学奖。投入占用产出技术是我国学者提出的。申请者曾应用此方法研究大规模投资和出口变动对中国经济影响，发现如不考虑乘数的时滞性，就往往误解为有关政策、措施等的效应都在当年发生，从而导致错误。此项目除了研究完全需要系数及乘数的时滞特征，时滞乘数及动态逆计算方法外，并建议研究具有多年时滞特性的新颖动态投入占用产出模型，如考虑人力资本多年时滞特性的动态模型、考虑自然资源勘探、准备和使用之间时滞动态模型、考虑科技投入与科技成果取得之间时滞动态模型等。由于连续型多年时滞动态投入占用产出模型的数学形式是一组微分差分方程，其结构系数矩阵为非线性，此模型的求解方法很困难，申请者提出利用阶变法进行求解。本项目成果将应用于中科院每年进行的GDP预测、全国粮食产量预测以及其他政策建议和分析报告中。

中文关键词： 投入占用产出技术；乘数效应；时滞；全国粮食产量预测；GDP增长速度预测

英文摘要： Input-output analysis is a widely useful approach in systems engineering and economic analysis.The approach creater, Wassily Leontief, is the winner of 1973 Nobel Prize in economic science. Input-occupancy-output technique was proposed by Chinese scholar and was highly appraised by some world well-known scholars. When we used the technique to study the effects of large scale investment and decrease of exports on Chinese economy, we find that if we don't use time-lag multiplier,we will make the following mistake that these policy effects will be taken in current year. In this proposal, we suggest to study not only time-lag feature of total requirement coefficient matrix, time-lag multiplier effects and dynamic inverse, but also multi-years time lag dynamic input-occupancy-output model, such as human capital multi-years time lag dynamic model, natural resource exploration and its use time-lag dynamic model, science and technological inputs and their achievements multi-years time lag dynamic model etc. Because the mathematical form of the continuous multi-years time lag dynamic input-occupancy-output model is a system of difference differential equations, its structural coefficient matrices are in non-linear form, and it is difficult to find the solution of equations. We propose to use step-variable methods to find

英文关键词： Input-occupancy-output technique；Multiplier effects；Time lag；National grain output prediction；GDP growth rate prediction