The distributed no-idle permutation flowshop scheduling problem with due windows algorithms

"The Distributed No-Idle Flowshop Scheduling Problem with Due Windows (DNIFSPDW)" addresses an extension of the Distributed Permutation Flowshop Scheduling Problem with No-Idle and Due Window constraints. The objective of DNIFSPDW is to determine the optimal sequence of job assignments to factories and the sequence in which they should be performed in each factory. This optimal sequence should ensure the minimum total weighted earliness and tardiness (TWET) penalties while also taking into account the due windows. The inclusion of a total weighted tardiness objective in the flowshop scheduling problem (FSP), a known NP-hard issue, means that the DNIPFSPDW also inherits this computational complexity, classifying it under the NP-hard category. As a result, exact solution methods are not efficient for large-scale instances of this problem. For such intricate challenges, metaheuristic approaches are more appropriate as they can achieve high-quality solutions in reasonable computation times. Notably, the iterated greedy metaheuristic has proven effective for most PFSPs. Therefore, two hybrid iterated greedy algorithms namely hybrid iterated greedy-tabu search and hybrid iterated greedy-local search are developed for the DNIPFSPDW.
The two attached documents contain the codes for the algorithms discussed. Each file includes the data developed and used to evaluate the distributed no-idle flowshop scheduling problem with due windows. For detailed instructions on executing these codes, please refer to the enclosed Readme file. The document titled "Java Codes" houses the executable codes.

带交货窗口的分布式无闲置流水车间调度问题（Distributed No-Idle Flowshop Scheduling Problem with Due Windows, DNIFSPDW）针对带无闲置约束与交货窗口约束的分布式排列流水车间调度问题进行了拓展。该问题的优化目标为确定工件向工厂分配的最优顺序，以及各工厂内工件的加工顺序，该最优调度方案需在兼顾交货窗口的前提下，实现总加权提前与拖期惩罚（total weighted earliness and tardiness, TWET）的最小化。流水车间调度问题（Flowshop Scheduling Problem, FSP）中引入总加权拖期惩罚目标属于已被证实的NP难问题，因此DNIFSPDW同样继承了该计算复杂度，归属于NP难范畴。针对该问题的大规模实例，精确求解方法的效率往往较低。对于这类复杂调度挑战，元启发式算法更为适用，其可在合理的计算时间内获得高质量的解决方案。值得注意的是，迭代贪心（iterated greedy）元启发式算法已被证实对多数排列流水车间调度问题（Permutation Flowshop Scheduling Problem, PFSP）有效。因此，本文针对DNIFSPDW开发了两种混合迭代贪心算法，分别为混合迭代贪心-禁忌搜索（hybrid iterated greedy-tabu search）算法与混合迭代贪心-局部搜索（hybrid iterated greedy-local search）算法。所附的两份文档包含了上述讨论算法的代码。每份文件均包含了用于评估带交货窗口的分布式无闲置流水车间调度问题的测试数据集。关于代码运行的详细操作说明，请参阅随附的Readme文件。标题为"Java Codes"的文档包含可执行代码。