采用主目标进化遗传算法的织造排程研究  

Research on weaving scheduling using main objective evolutionary genetic algorithm

作　　者：孟朔 潘如如[1] 高卫东[1] 王静安 周利军 MENG Shuo;PAN Ruru;GAO Weidong;WANG Jing′an;ZHOU Lijun(Key Laboratory of Eco-Textiles (Jiangnan University),Ministry of Education,Wuxi,Jiangsu 214122,China;Danyang Dansheng Textile Co. ,Ltd. ,Zhenjiang,Jiangsu 212309,China)

机构地区：[1]生态纺织教育部重点实验室(江南大学),江苏无锡214122 [2]丹阳市丹盛纺织有限公司,江苏镇江212309

出　　处：《纺织学报》2019年第8期169-174,共6页Journal of Textile Research

基　　金：国家重点研发计划项目(2017YFB0309200);中央高校基本科研业务费专项资金项目(JUSRP51907A)

摘　　要：针对目前织造企业在订单多、批量小、品种杂、交期严格的情况下,织造排程效果差的问题,提出一种采用主目标进化遗传算法的织造排程方法。通过建立一个由织机类型约束的并行织造排程模型,以满足订单交期为主目标,最小化品种翻改、完工时间等为副目标,改进了普通快速非支配遗传算法,并以某织造车间的生产实例进行仿真实验。结果表明,此主目标进化算法排程效果优于企业人工排程方法,能够指导实际生产。为提高算法的鲁棒性,利用自适应调整遗传参数对算法优化,进一步提高质量,且在一定程度上避免了早熟,增强全局搜索能力。In order to solve the problems of large quantities, small specifications of orders, numerous products and strict delivery requirements in the textile industry, an automatic weaving scheduling method using main evolutionary genetic algorithm was proposed. A parallel weaving scheduling model with different loom type constraints was established to modify a common fast non-dominant genetic algorithm with the main goal of meeting the order delivery date and secondary goal of minimizing the change of variety and minimizing the completion time. Then a simulation experiment of the scheduling scheme was carried out. The result show that the effect of each objective function superior to the manual scheduling method, which shows a certain practical application effect to solve the current weaving scheduling problem. In order to further improve the performance of the algorithm, an adaptive method was carried out to optimize the genetic parameters. The optimized algorithm can avoid the precocity and enhance the global search ability, and the objective function of optimized algorithm is further improved.

关 键 词：织造 生产排程 多目标优化 非支配遗传算法 

分 类 号：TS108.8[轻工技术与工程—纺织工程]