时间-费用相关的随机活动网络仿真研究  

Simulation Research of Stochastic Activity Network with Correlative Variables

作　　者：杨宝森[1,2] 刘鲁[1] 马晨[3] 

机构地区：[1]北京航空航天大学经济管理学院,北京100083 [2]天津商业大学,天津300134 [3]中国农业大学,北京100083

出　　处：《系统仿真学报》2010年第4期864-866,共3页Journal of System Simulation

基　　金：国家自然科学基金(70671007);教育部博士点基金课题(200800060005)

摘　　要：针对复杂产品研发项目活动时间、费用不能准确确定,采用蒙特卡罗仿真方法,建立了时间和费用为相关随机变量的随机活动网络仿真模型。利用乔列斯基因子分解法,将独立分布的时间和费用随机变量转化为相关的随机变量,通过对仿真结果的分析,可以更准确地进行时间-费用交换分析,优化项目工期和费用。通过算例说明了仿真建模和分析过程。Since complex product development project is characteristic of uncertain activity time and cost,the Monte Carlo simulation method can be used to model the stochastic activity network with the correlative stochastic variables of time and cost. Here the method of Cholesky decomposition has been used to convert independent distribution of stochastic variables such as time and cost into correlative stochastic variables. The simulation can accurately and rapidly process the job of time-cost trade-off and obtain optimum project duration and cost. At last,a case study illustrates the modeling and analysis process.

关 键 词：仿真 随机活动网络 时间-费用交换 蒙特卡罗 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术] N945.13[自动化与计算机技术—计算机科学与技术]