A simulation-based differential evolution algorithm for stochastic parallel machine scheduling with operational considerations
A simulation-based differential evolution algorithm for stochastic parallel machine scheduling with operational considerations
复制标题
DOI：
10.1111/itor.12011
复制
发表时间：
2013-07
期刊：
影响因子：
--
通讯作者：
Rui Zhang;Shiji Song;Cheng Wu
中科院分区：
文献类型：
--
作者：
Rui Zhang;Shiji Song;Cheng Wu
我们考虑了一个平行机调度问题，目标是最小化两类成本：与生产操作相关的成本和与交货期性能相关的成本。前者可以通过合理设置操作变量(例如工人数量、维护频率)来减少，而后者可以通过对生产过程进行适当的调度来减少。然而，由于人为因素的影响，这两个目标的优化都变得非常复杂，而人为因素在现实世界的制造系统中发挥着主导作用。针对这一问题，本文采用了一种基于仿真的优化框架来获得集成调度问题的高质量鲁棒解。同时，采用一种基于群体智能的元启发式算法--差异进化算法，对庞大的解空间进行系统搜索。最后，通过数值计算验证了该方法的有效性。文中还给出了灵敏度分析和实际意义。
We consider a parallel machine scheduling problem with the objective of minimizing two types of costs: the cost related to production operations and the cost related to due date performances. The former could be reduced by reasonable settings of the operational variables (e.g., the number of workers, the frequency of maintenance), while the latter could be reduced by appropriate scheduling of the production process. However, the optimization of both targets is significantly complicated by the influence of human factors that play a dominant role in real-world manufacturing systems. To cope with this issue, a simulation-based optimization framework is adopted in this paper for obtaining high-quality robust solutions to the integrated scheduling problem. Meanwhile, differential evolution, a metaheuristic algorithm based on swarm intelligence, is applied for a systematic search of the huge solution space. Finally, numerical computations are conducted to verify the effectiveness of the proposed approach. Sensitivity analysis and practical implications are also presented.