引用本文:万晓琴,严洪森.面向航空发动机装配线的知识化制造系统重调度和自重构[J].控制理论与应用,2017,34(7):903~911.[点击复制]
WAN Xiao-qin,YAN Hong-sen.Rescheduling and self-reconfiguration of knowledgeable manufacturing system oriented to aircraft engine assembly line[J].Control Theory and Technology,2017,34(7):903~911.[点击复制]
面向航空发动机装配线的知识化制造系统重调度和自重构
Rescheduling and self-reconfiguration of knowledgeable manufacturing system oriented to aircraft engine assembly line
摘要点击 2289  全文点击 1868  投稿时间:2016-10-28  修订日期:2017-03-29
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2017.60812
  2017,34(7):903-911
中文关键词  知识化制造  自重构  装配  重调度
英文关键词  knowledgeable manufacturing  self-reconfiguration  assembly  rescheduling
基金项目  (61673112, 60934008), 东南大学优秀博士论文(YBJJ1446), 中央高校基本科研业务费专项资金(2242014K10031), 江苏高校优势学科建设工程资助项目资助.
作者单位E-mail
万晓琴* 东南大学 自动化学院 xqwan87@163.com 
严洪森 东南大学 复杂工程系统测量与控制教育部重点实验室  
中文摘要
      为了解决航空发动机装配过程中存在的不确定返工问题, 建立了面向航空发动机装配线的知识化制造系统(KMS)重调度和班组自重构优化模型, 提出装配线重调度及自重构集成优化算法. 在算法调度层, 证明了以加权完工成本为优化目标的工序排序性质, 并对工序进行初始排序. 定义了三种邻域结构, 用变邻域搜索(VNS)对工序在并行装配组上的指派问题和调度问题进行优化. 在重构层, 在不违背装配组装配技能约束的前提下利用装配线负载平衡原则对装配班组进行配置, 并采用禁忌搜索(TS)对班组配置进行优化. 仿真实验结果表明了模型与算法的有效性.
英文摘要
      To solve the assembly probem of the aircraft engine with unceratin rework, an optimization model of recheduling and teams self-reconfiguration of knowledgeable manufacturing system oriented to aircraft engines assembly line is build, and an integrated optimization algorithm of both is proposed. InStheSscheduling phase, dominance relations of operations aiming at optimizing weighted completion cost are derived and applied to generating an initial operation sequence. Three neighbourhood structures are defined. A variable neighbourhood search is used to optimize the the assignment and the schedule of operations on parallel assembly teams. InSthe self-reconfigurationSphase, the configuration of the assembly teams is generated according to the assembly line workload balancing without violating the assembly skills constraints, and then optimized by a tabu search. The simulation experiments validate the effectiveness of the proposed model and algorithm.