In project management, a project plan is constructed that assigns a planned start time to each project activity. Based on this plan, the total planned project duration and cost can be determined. However, during project execution, deviations from the plan are inevitable due to uncertainty and variability. When these deviations endanger the timely completion of projects, the project manager should take corrective actions to get the project back on track. In this study, corrective actions are modelled as modifications of the original activity duration distributions (i.e., reduced mean and/or standard deviation) to account for the uncertain nature of their impact. Further, an analytical procedure is developed to rank activities according to their expected impact on the project duration distribution when they are controlled by a corrective action. This activity ranking is used to determine the number of actions that should be taken and to select the set of activities that will be controlled. A computational experiment on a large set of project networks with varying network complexity and network structures has been conducted. These experiments have shown that taking actions on a relatively small subset of activities, rather than on the entire set of project activities, proves more efficient, when the subset of activities is carefully selected. More precisely, the efficiency of the corrective actions process depends on both the number of actions and the activity selection criterion (activity ranking).

中文翻译：

---

随机项目网络纠正措施的影响分析

在项目管理中，构建项目计划，为每个项目活动分配计划的开始时间。根据该计划，可以确定计划项目的总工期和成本。然而，在项目执行过程中，由于不确定性和可变性，与计划的偏差是不可避免的。当这些偏差危及项目的及时完成时，项目经理应该采取纠正措施，使项目回到正轨。在本研究中，纠正措施被建模为对原始活动持续时间分布的修改（即减少平均值和/或标准差），以解释其影响的不确定性。此外，还开发了一种分析程序，根据活动受纠正措施控制时对项目工期分布的预期影响对活动进行排名。该活动排名用于确定应采取的操作数量并选择将受到控制的活动集。对大量具有不同网络复杂度和网络结构的项目网络进行了计算实验。这些实验表明，当仔细选择活动子集时，对相对较小的活动子集而不是对整个项目活动集采取行动会更有效。更准确地说，纠正措施过程的效率取决于行动的数量和活动选择标准（活动排名）。