Abstract: During the cutting process of metal plate,the jobs need to be scheduled on many different kinds of machines with the same function.The studied problem belongs to the unrelated parallel machine scheduling problem which has a great influence on the manufacturing cost of the product directly.The cost components of the metal plate cutting process were analyzed,the constraints of the metal plate cutting scheduling problem were summarized.Then,the mathematical model of the multi-constraint complex optimization problem was established with the total production cost as an objective function.Due to the NP-hard attribution,an improved decimal grey wolf optimization algorithm was proposed.Decimal integer coding was introduced by the position of artificial wolves,and the intelligent behaviors such as scouting and running move operators were redesigned.The proposed algorithm was tested by multiple instances and compared with the Genetic Algorithm(GA),the Harmony Search Algorithm(HSA),the experimental results showed that the proposed algorithm was both feasible and efficient.

Key words: metal plate, cutting process, grey wolf algorithm, parallel machine scheduling, total production costs

摘要： 金属板材在切割下料工序中,需要在多台不同类型的设备上进行调度,属于不相关并行机调度问题,调度的优良程度直接影响制造成本。分析了金属板材切割工序中的成本构成,总结了板材切割调度问题的约束条件,以总生产成本为目标函数,建立了多约束复杂并行机调度数学模型。由于该问题是NP-hard问题,提出一种十进制灰狼优化算法进行求解。对人工狼的位置进行十进制整数编码,重新设计游走和奔袭等智能行为。用多组算例对所提求解算法进行测试,同时与遗传算法以及和声算法进行对比,结果表明了所提算法的实用性和有效性。

关键词: 金属板材, 切割工序, 灰狼算法, 并行机调度, 总生产成本