特殊工艺约束下最小完工时间并行多机调度问题的研究

Research on parallel multi-machine scheduling problem for minimum makespan under special process constraint

非等同并行机最小化完工时间调度问题作为家纺企业车间调度问题的重要组成部分,有着独特的特点,一方面并行多机非等同,另一方面每机器可生产产品类型受特殊工艺的约束,针对该问题的特点,基于免疫系统的克隆选择原理,结合一种新型的促进和激励群体多样性的技术,提出了一个新颖的人工免疫算法.仿真结果表明,此算法是有效的,优于遗传算法和克隆选择算法,并能适用于解实际家纺企业这类调度问题.

Non-equivalent parallel multi-machine scheduling problem for minimizing the makespan in textile enterprises is an important part of job-shop scheduling problem. It has some special characteristics. On one hand, parallel multi-machines is non-quivalent. On the other hand, the kinds of products processed on each machine are subject to special process constraint. In order to solve this kind of scheduling problem in textile enterprises, a novel artificial immune algorithm based on clonal selection theory is presented. The proposed algorithm composes of a new technique that is used to promot and stimulate diversity of the population. The computational results show that this algorithm is effective and is better than genetic algorithm and clone selection algorithm. It can be applied to solve practical scheduling problems for minimizing the makespan with special process constraint in textile enterprises.