Abstract: Resource leveling is a classic scheduling problem in repetitive projects. This paper presents a method of resource leveling for repetitive projects in line of balance (line of balance, LOB). Firstly, it proposes a method to determine the critical path in LOB, and determines the critical path and critical activities types. Then, it analyzes the determinants of the project duration, and studies the characteristics of different types of critical activities, and the relationship between the total duration and resource adjustment. It proves that due to the existence of special activities, namely, the backward critical activity and the point critical activity, resource optimization can be achieved by adjusting critical activities and non-critical activities simultaneously under the premise of ensuring the project duration. According to this idea, the paper designs the genetic algorithm for resource leveling in LOB. Finally, an illustrative pipeline project is used to demonstrates the optimization performance of resource leveling model. The resource leveling ideas and algorithms proposed in the paper can help project planners to expand the resource optimization space and achieve better resource leveling effect.

Key words: repetitive project, line of balance, critical activity, resource leveling

摘要： 资源均衡是重复性项目中的经典调度问题,本文提出一种新的基于平衡线法(line of balance,LOB)的资源均衡方法。首先,本文提出LOB中关键路线的确定方法,确定关键路线及关键工序类型。而后,本文分析项目总工期的决定因素,对不同类型关键工序的特性及其与总工期、资源调整之间的关系进行了研究,论证了在LOB的资源均衡问题中,由于逆关键工序、点关键工序这些特殊工序的存在,可以在保证项目总工期不变的前提下,通过同时调整关键工序和非关键工序实现资源优化。按照这一思路,论文设计了LOB中资源均衡的遗传算法。算例分析表明该资源均衡算法的优化性能。本文提出的资源均衡思路和算法能帮助项目计划人员拓展资源优化空间,达到更好的资源均衡效果。

关键词: 重复性项目, 平衡线法, 关键工序, 资源均衡