基于机动时间的可分解平行工序顺序优化研究

The Sequencing Optimal Decision with Divisible Activities

平行工序的顺序优化是解决资源有限项目进度计划问题的最有效、最普遍的方法之一。对于该类问题的研究目前主要基于工序的不可分解性,而现实情况下有些工序是任意可分的。基于此,本文首先提出了最小路长定理,在其基础上,建立了任意可分的两个平行工序调整为顺序工序的亏值模型,并进行了理论证明,此外,针对从n个可分解平行工序中选取一个与指定工序调整为顺序工序的优化问题进行了研究,在已给亏值模型的基础上设计出了优化算法,越是大型网络,该方法的优越性越明显。

The sequencing optimal decision of parallel activities is one of the most effective and familiar methods in the issue of resource--constrained schedule planning. Researches on this problem have focused on the indivisible activities mainly ever before. In fact, many activities can be decomposed into arbitrary portions. In this case, this paper proposed the minimum project duration theory and on the foundation established the tardiness model of sequencing the two divisible parallel activities, then gave the theoretic proof. Furthermore, the problem of sequencing an activity selected among n divisible parallel activities with a designated activity is also studied in this paper. Based on the tardiness model, the optimal method and theoretic proof are provided. Especially for a large scale network, the superiority becomes very distinct.