项目的工序间存在着普遍的相依性,在项目的进度管理中,合理考虑工时之间的相依性,有利于更加准确地预测和评估工期风险,从而更有效地控制工程进度.本文建立了一个基于贝叶斯网络的进度计划动态更新模型,度量工程进度网络中工序持续时间的相依随机性,并进行进度计划的动态更新,模型用贝叶斯网络表示工序间的相依关系,由专家估计工序持续时间的边缘分布及工序间协调系数,然后确定贝叶斯网络中的条件概率和分布,从而确定各工序持续时间的条件分布和总工期的分布.算例表明该模型能有效预测和控制工时的不确定性,有利于降低工程进度风险.%There is the universal dependence between the project processes. In the project schedule management, considering the dependence between time reasonably, is conducive to predicting and assessing the schedule risk more accurately, so as to effectively control the progress of the project. This paper establishes an dynamic updating model of the schedule based on Bayesian network, which is used to measure the dependent random of the process duration in the project progress network, and dynamically update the schedule. The model uses Bayesian network to represent dependence between processes. The marginal distribution of process duration and the coordination coefficient between processes are estimated by experts, and then the conditional probability and distribution in the Bayesian network are determined, so as to determine the conditional distribution of each process duration and total duration. The example shows that the model can effectively predict and control the uncertainty of the working hours, which is conductive to reducing the project risk.
展开▼
