工业工程

工业工程 ›› 2023, Vol. 26 ›› Issue (5): 89-96,114.doi: 10.3969/j.issn.1007-7375.2023.05.010

• 系统建模与优化算法 • 上一篇    下一篇

多技能资源能力不均衡环境下项目调度的鲁棒优化方法

胡振涛1, 崔南方2   

  1. 1. 湖北经济学院 工商管理学院，湖北 武汉 430205;
    2. 华中科技大学 管理学院，湖北 武汉 430074
  • 收稿日期:2022-06-02 发布日期:2023-10-25
  • 通讯作者: 崔南方(1963-),江西省人,教授,博士,主要研究方向为项目调度、供应链管理。E-mail:nfcui@mail.hust.edu.cn E-mail:nfcui@mail.hust.edu.cn
  • 作者简介:胡振涛(1988-),河南省人,讲师,博士,主要研究方向为项目调度、组合优化
  • 基金资助:
    国家自然科学基金资助项目 (71971094，71701067)

Robust Optimization for Multi-skilled Project Scheduling with Uneven Resource Capacities

HU Zhentao1, CUI Nanfang2   

  1. 1. School of Business Administration, Hubei University of Economics, Wuhan 430205, China;
    2. School of Management, Huazhong University of Science and Technology, Wuhan 430074, China
  • Received:2022-06-02 Published:2023-10-25

RichHTML

351

PDF

896

摘要： 现实项目在实施过程中面临着诸多不确定因素，鲁棒项目调度是应对项目不确定性，减少进度偏差的有效手段。此外，项目中还广泛存在一类能力不均衡的多技能资源，这类资源会提高调度计划制定的难度，但同时资源之间更为灵活的替代、合作关系也为鲁棒项目调度提供了更大的优化空间。基于此，设计两阶段算法求解不确定环境下含有此类资源的项目鲁棒调度计划。第1阶段构造基于规则的启发式算法：结合活动优先规则、资源权重规则，通过0-1规划模型求解基准调度计划及资源分配方案。第2阶段设计鲁棒优化算法：通过时间缓冲的有偏随机插入与回退，以及资源分配方案的调整对基准调度计划进行鲁棒优化。仿真实验表明，不同风险水平下，算法在求解不同规模的项目算例时，所得的调度计划在鲁棒性方面均表现出了明显的优势。

关键词: 多技能资源, 能力不均衡, 项目调度, 鲁棒优化, 分散缓冲

Abstract: There are many uncertain factors during the implementation of real-life projects. Robust scheduling is an effective method to deal with uncertainties and reduce schedule deviations of a project. In addition, the multi-skilled resources with uneven capacities widely exist in real projects, which may increase the difficulty of scheduling. However, such resources can also enlarge the optimization space for robust project scheduling due to the flexible substitution and cooperation relationships among them. Based on this, a two-stage algorithm is proposed to solve the multi-skilled project scheduling problem with uneven resource capacities (URC-MSPSP). In the first stage, a rule-based heuristic algorithm is designed through combining activity priority rules and resource weight rules. A 0-1 linear programming model is built for solving the baseline scheduling plan and resource allocation strategies. Then in the second stage, a robust optimization algorithm is designed for the baseline scheduling plan by biased random insertion and deletion of time buffering, as well as the adjustment of resource allocation strategies.. Simulation experiments show that the proposed algorithm is significantly superior to other algorithms in terms of robustness for projects with different scales under different risk levels.

Key words: multi-skilled resource, uneven capacity, project scheduling, robust optimization, scattered buffer

中图分类号: