Abstract: In order to enhance management efficiency, Goldratt proposed the critical chain project management method in 1997. This method applies the theory of constraints, which was initially used in the production field, to the realm of project management. Instead of focusing on the critical path, this method highlights the critical chain. Additionally, it places a greater emphasis on the limitation of resource constraints. To mitigate the impact of uncertainty, the method incorporates a buffer into the project timeline. This not only shortens the project duration but also reduces the impact of human behavioral factors on the project. Simultaneously, the project is monitored and controlled through reasonable measures during its implementation.
Classical research presumes that buffers can be obtained at any time. However, it is often difficult to achieve in the practical implementation of projects. Excessive monitoring can result in project deviation from the initial plan. To ensure that each buffer monitoring is beneficial for the project implementation, this study proposes a method for determining the monitoring period based on project objectives and activity risks. Different projects have varying priorities regarding schedule, cost, and quality. The relationship between these factors is utilized to determine the appropriate monitoring frequency for different project priorities. The monitoring interval for each activity is then allocated according to the risk level of each activity. This determines the monitoring period for each activity. Consequently, the project can attain optimal benefits with a limited number of buffer monitoring instances.
The purpose of monitoring buffer consumption is to ensure controllability during project execution and make timely adjustments when deviations occur. The classical approach to monitoring critical chain projects considers the impact of monitoring triggers and makes buffer monitoring a continuous state by default. However, frequent monitoring consumes considerable manpower and time, increasing project costs. Therefore, setting a reasonable monitoring period and frequency is crucial during actual project monitoring. The main objective in setting the monitoring cycle is optimising project comprehensive benefits. As various uncertainties affect project comprehensive benefits during implementation, the monitoring interval should be adjusted according to project objectives and activity risks. A uniform monitoring cycle should not be used to achieve optimal monitoring effects with appropriate efforts. The monitoring cycle should not be standardised to achieve optimal results with the most appropriate monitoring efforts.
Each project differs in duration, cost, and quality requirements. These factors interact, and shortening duration increases costs and decreases quality. Improving quality requires extending the duration and increasing costs, but high-quality projects reduce rework costs later. Therefore, ensuring project comprehensive benefits is optimal. We determine the overall monitoring frequency based on project objectives, and then analyse activity risk impacts to set the monitoring period. In the planning stage, we estimate multiple risk impacts for each project activity. In the execution stage, we calculate the buffer consumption risk impacts for each activity by determining the ratio of various buffer usage to the planned buffer amount. We analyse the comprehensive risk impacts of activities using the critic method and calculate the comprehensive risk impact coefficients. We combine these with the monitoring frequency to set a reasonable monitoring cycle.
Monte Carlo simulation is made using Matlab, assuming each activity duration obeys the lognormal distribution, and generates 1000 random durations using Matlab. The experimental simulation using the static three-part method, the relative buffer monitoring method, and the proposed method shows that, compared to the comparative methods, the proposed method can shorten project duration and reduce costs while meeting quality requirements.

Key words: project management, critical chain, project objectives, monitoring cycle, activity risk

摘要： 经典研究假设随时能够得到缓冲的消耗状态,但在项目实际执行的过程中很难实现,且监控次数过多会导致项目偏离计划。为保证每一次缓冲监控都能够有利于项目实施,本文提出了一种基于项目目标和活动风险确定监控周期的方法。不同项目对工期、成本、质量的侧重点不同,利用工期与成本、工期与质量之间的关系,根据不同的侧重点确定合适的监控频次。再根据每项活动的风险大小分配监控次数,活动的风险越大需分配到的监控次数越多,从而确定每项活动的监控周期,使项目在有限次的缓冲监控下实现效益最优。最后利用Matlab进行蒙特卡洛模拟实验,结果表明,与对比方法相比,所提方法能在满足项目质量要求的前提下,缩短项目工期并降低成本。

关键词: 项目管理, 关键链, 项目目标, 监控周期, 活动风险