基于贝叶斯网络的工程项目进度完工概率分析

针对传统计划评审技术(Program Evaluation and Review Technique,PERT)在计算完工概率时假设条件的局限性(假设条件与工程实际存在偏差,导致完工概率偏大),提出了基于贝叶斯网络的施工进度完工概率分析方法.首先,分析了贝叶斯网络与进度计划网络之间的相似性,将两者结合起来构建了贝叶斯进度网络;在此基础上,综合考虑贝叶斯网络在节点取值及概率计算方面的优越性,并结合工程项目的不确定性及复杂性特点,建立了基于贝叶斯网络的施工进度完工概率分析模型.最后,将该模型应用于具体工程进行实例分析,验证了模型的可行性与有效性.研究结果表明:基于贝叶斯网络的进度完工概率模型充分考虑了工程施工中的风险因素,其结果能更客观地反映工程实际,可为工程项目决策者提供可靠的依据.

Analysis of Project Construction Schedule Completion Probability Based on Bayesian Network

In light of the limitation of hypothetical terms when using Program Evaluation and Review Technique (PERT) to calculate completion probability (Due to the deviation between hypothetical terms and practical engineering,resulting in larger completion probability),the analytical method of construction progress completion probability based on Bayesian network is put forward.Firstly,analyzing the similarity between Bayesian network and network schedule,making a combination of them to build a Bayesian schedule network.On this basis,comprehensively considering the superiority of nodes value and probability calculation in Bayesian network and combining uncertainty and complexity of the project itself,a model of construction progress completion probability based on Bayesian network is established.Finally,verifying its feasibility and effectiveness by applying it in practical project.The result shows that the model of construction progress completion probability based on Bayesian network has been fully considered risk factors resulting from engineering construction,furthermore,its result can objectively reflect the actual project,which proves that this method can provide a reliable evidence for engineering decision makers.