基于Monte Carlo多重仿真的费用与进度联合置信估计  

Confidence Percentile Estimation to Cost and Schedule Integration Based on Monte Carlo Multiple Simulation Analysis Technique

作　　者：徐哲[1] 吴瑾瑾[1] 汪阳青[1] 

机构地区：[1]北京航空航天大学经济管理学院,北京100083

出　　处：《系统仿真学报》2006年第12期3334-3337,共4页Journal of System Simulation

摘　　要：采用独立估计的方法不能同时获得高置信百分位的费用估计与进度估计,实际项目执行的费用风险或进度风险很高。费用与进度的联合置信估计是指费用置信百分位点估计和进度置信百分位点估计的联合估计。采用MonteCarlo多重仿真技术、回归分析与统计方法相结合的综合方法,在费用(进度)置信百分位估计的条件下,建立进度(费用)估计与相应的估计置信百分位之间的回归模型,由此获得费用(进度)置信百分位估计条件下的进度(费用)的置信百分位估计。Using independent estimating techniques, the high confidence percentile estimating values for cost and schedule can＇t be gained simultaneously, and the cost risk and the schedule risk are increasing in the actual projects. The integration method combines Monte Carlo multiple simulation analysis technique, regression analysis and statistical analysis. First, the project with stochastic cost and duration in activities was simulated m times using n runs per simulation. Then, the simulation outputs of cost and schedule were analyzed. The 95th percentile project cost （or schedule） value, its corresponding schedule （or cost） value, and the percentile ranking of the schedule （or cost） were recorded, and the conditional percentile ranking of the schedule （or cost） values were determined. Finally, in order to quantify the relationship existing between the percentile ranking of the schedule （or cost） and its corresponding schedule （or cost） value, the regression analysis models about them were founded and compared, and the high confidence percentile estimating values for cost and schedule were required simultaneously.

关 键 词：费用与进度 联合置信估计 置信百分位 MONTE Carlo多重仿真 回归分析 

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]