制造系统通用作业计划与蚁群算法优化

Universal shop scheduling problem for manufacturing system and ant colony optimization

分析了基本作业方式的"通用—特例"关系,提出了全面研究上述基本作业方式的通用作业计划的概念,建立了通用作业计划非连接图模型及基于蚁群算法的优化方法。算例研究假定一制造系统包含开放作业、异顺序作业、混杂流水作业三种作业方式,将其按照通用作业计划方式采用蚁群算法进行优化,所得的近优加工流程时间显著优于采用穷举法对三种作业方式分别优化所得的最优解。

After the relationship of Universal-Special in basic shops was analyzed, the concept of Universal Shop scheduling Problem (USP), which included all basic shop scheduling problems, was put forward. A disjunctive graph model representing USP was established, and the Ant Colony Optimization (ACO) used for USP was demonstrated. A case study assumed that a manufacturing system contained OSP, JSP and HFSP. When the system was treated as USP, the approximately optimal makespan computed by ACO was obviously better than the least makespan by enumerating all the possible solutions as the three shops were scheduled separately.