This paper studies the scheduling problem of jobs with release times, non-identical sizes, and incompatible job families on unrelated parallel batch machines. The capacities of batch machines and the processing times of each job on the batch machines are different. The processing time of one batch is equal to the longest processing time of jobs in this batch. Different types of jobs are not allowed to be assigned into the same batch, which is known as incompatible job families. Mixed integer linear programming and constraint programming (CP) models are proposed. A new batch-based local search method is designed and an iterated greedy (IG) algorithm is developed to avoid unreasonable exchanging of jobs during the local search. Numerical results show that the CP method can obtain high quality solutions in the small-scale instances. For the large-scale instances, the IG algorithm with the new local search method has a competitive performance.

本文研究了在不相关的并行批处理计算机上具有发布时间，大小不相同以及作业系列不兼容的作业的调度问题。批处理机的容量和批处理机上每个作业的处理时间是不同的。一批的处理时间等于该批中最长的作业处理时间。不允许将不同类型的作业分配到同一批次中，这称为不兼容的作业族。提出了混合整数线性规划和约束规划（CP）模型。设计了一种新的基于批处理的本地搜索方法，并开发了迭代贪婪（IG）算法，以避免在本地搜索过程中不合理地交换工作。数值结果表明，CP方法可以在小规模实例中获得高质量的解。