This paper deals with the no-wait multiproduct multistage product scheduling problem (NWMMSP) with minimum makespan and maximum mean customer’s satisfaction level criteria. The mean customer’s satisfaction level criterion related to flexible due date is introduced to approach an imprecise or flexible nature of the data in a practical manufacturing environment. A novel discrete multi-objective evolutionary algorithm combined with an improved differential evolution algorithm and memetic algorithm (MOIDE-MA) is proposed to solve the multi-objective NWMMSP. The improved discrete differential evolution algorithm (IDE) mainly focuses on the edge area of the Pareto front (PF) space and the single objective guide strategy to construct better reference points. Meanwhile, an information-sharing strategy is introduced in IDE to improve the information diversity and enhance the search capabilities for the whole PF. A novel memetic algorithm (MA) based on genetic algorithm (GA) and local search related to the researched scheduling problem is developed to improve the distribution uniformity and accuracy of the non-dominated solutions obtained by IDE further. IDE gains some useful information extracted from the external archive of MA to speed up its search. Computational and simulation results show that the improvement strategy proposed in the algorithm can effectively improve the performance of MOIDE-MA. Furthermore, the proposed algorithm is compared with three other multi-objective scheduling optimization approaches based on different scale multi-objective NWMMSP instances. The experimental results show the effectiveness of the proposed MOIDE-MA.

本文处理具有最小制造跨度和最大平均客户满意度标准的无等待多产品多阶段产品调度问题 (NWMMSP)。引入了与灵活到期日相关的平均客户满意度标准，以解决实际制造环境中数据的不精确或灵活性质。提出了一种新的离散多目标进化算法，结合改进的差分进化算法和模因算法（MOIDE-MA）来解决多目标NWMMSP。改进的离散差分进化算法（IDE）主要针对帕累托前沿（PF）空间的边缘区域和单目标引导策略来构建更好的参考点。同时，在IDE中引入了信息共享策略，以提高信息的多样性并增强整个PF的搜索能力。针对所研究的调度问题，开发了一种基于遗传算法 (GA) 和局部搜索的新型模因算法 (MA)，以进一步提高 IDE 获得的非支配解的分布均匀性和准确性。IDE 从 MA 的外部档案中提取一些有用的信息以加快其搜索速度。计算和仿真结果表明，算法中提出的改进策略能够有效提高MOIDE-MA的性能。此外，将所提出的算法与基于不同规模多目标 NWMMSP 实例的其他三种多目标调度优化方法进行了比较。