基于多群体协同进化混合算法的FJSP研究

发布时间：2018-01-08 16:11

本文关键词：基于多群体协同进化混合算法的FJSP研究　出处：《组合机床与自动化加工技术》2017年01期 　论文类型：期刊论文

【摘要】：为更有效地解决柔性作业车间调度问题,提出一种多群体协同进化混合算法,该混合算法的主群与子群分别采用带有精英保留策略的遗传算法与粒子群算法。算法初期主群与子群以不同的策略独立并行地进行寻优,后期主群与子群、子群之间按照一定的规则进行信息的交流以实现协同进化,从而提高算法前期全局搜索能力与后期局部挖掘能力。此外,借助信息熵实现了主群交叉、变异概率的自适应调整并运用凹函数递减策略对三个子群的惯性权重值进行动态调整以提高该混合算法的整体优化性能。最后,通过Kacem基准问题验证了该算法求解柔性作业车间调度问题的有效性。
[Abstract]:In order to solve the flexible job shop scheduling problem more effectively, a multi-group co-evolutionary hybrid algorithm is proposed. The main group and subgroup of the hybrid algorithm adopt the genetic algorithm and particle swarm optimization algorithm with elite reservation strategy respectively. In the initial stage of the algorithm, the principal group and the subgroup are independently and paralleled with different strategies to optimize, and later the principal group and the sub-group. In order to realize the co-evolution, the sub-groups exchange information according to certain rules, so as to improve the ability of global search and local mining in the early stage of the algorithm. In addition, the information entropy is used to realize the intersection of the main groups. Adaptive adjustment of mutation probability and dynamic adjustment of inertial weight of three subgroups by decreasing concave function strategy to improve the overall optimization performance of the hybrid algorithm. The effectiveness of the proposed algorithm for flexible job shop scheduling problems is verified by the Kacem benchmark problem.
【作者单位】：中航工业成都飞机工业(集团)有限责任公司;西北工业大学管理学院;
【基金】：国家自然科学基金项目(U1404702)
【分类号】：TH165;TP18
【正文快照】： 710129)0引言进入21世纪以来,新一轮科技革命和产业变革蓄势待发,中国正在从“制造大国”向“制造强国”迈进。《中国制造2025》明确指出“以推进智能制造为主攻方向”,智能制造主要体现在“动态感知、实时分析、自主决策、精准执行”四个方面的特征,而其中“自主决策”是智能

【相似文献】