传统的计划评审技术(PERT)考虑的三时估计值主观性偏强,没有考虑风险因素相关性对工序活动工期的影响,影响了工期仿真与风险评价的客观性以及分析结果的可靠性。本文建立了耦合改进PERT和贝叶斯信念网络(BBNs)的心墙堆石坝施工进度风险分析模型,首先利用BBNs推理功能推算工序活动风险因素相关性,同时分析工程进度风险;其次将三角模糊数和模糊测度Me方法相结合来对PERT进行改进,引入了决策者风险偏好指标;最后将该模型应用到实际工程,得到决策者风险偏好指数-工期-完工概率分布曲线,可用于指导施工决策。工程应用结果表明:与传统的PERT方法相比,耦合改进PERT和BBNs得到的仿真结果更加可靠,可为心墙堆石坝施工进度方案设计提供理论依据。
Traditional project evaluation and review technique(PERT) considers the three-time estimateswhose subjectivities are too strong,and doesn't consider the effect that the correlations of risk factors imple-menting on the project duration,influencing the objectivity of risk assessment and the reliability of analysisresults. This study proposes the coupled improved PERT-Bayesian belief networks(BBNs) model. First,the reasoning function of BBNs is used to analyze the correlations among risk factors. Second, combiningthe method of Triangular Fuzzy Number and Fuzzy Measure to optimize PERT and introducing the risk pref-erence index of decision makers. Then apply the model to practical engineering, obtaining the risk prefer-ence index-period-completion probability distribution curve,which can be used to provide guidance for deci-sions. An engineering application result shows that compared with the traditional PERT,the method obtainsmore reliable simulation results and provide a reliable theoretical basis for actual construction schedule de-signing.