引用本文:吴子轩,李铁克,张文新,王柏琳.考虑机器检修的热轧钢管批量计划方法[J].控制理论与应用,2017,34(9):1250~1259.[点击复制]
WU Zi-xuan,LI Tie-ke,ZHANG Wen-xin,WANG Bai-lin.Methods of hot-rolled batch planning for seamless steel tube with machine maintenance[J].Control Theory and Technology,2017,34(9):1250~1259.[点击复制]
考虑机器检修的热轧钢管批量计划方法
Methods of hot-rolled batch planning for seamless steel tube with machine maintenance
摘要点击 2180  全文点击 1235  投稿时间：2017-03-06  修订日期：2017-06-08
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2017.70131
  2017,34(9):1250-1259
中文关键词  无缝钢管  热轧批量计划  机器检修  机器调整时间  启发式算法
英文关键词  seamless steel tube  hot-rolled batch planning  machine maintenance  setup times  heuristic algorithm
基金项目  国家自然科学基金项目(71701016,71231001), 中央高校基本科研业务费项目(FRF–BD–16–006A), 北京市自然科学基金项目(9174038), 教育部人 文社会科学研究青年基金项目(17YJC630143)资助.
作者单位E-mail
吴子轩 北京科技大学 东凌经济管理学院 zixuan_wu@sina.com 
李铁克* 北京科技大学 东凌经济管理学院  
张文新 北京科技大学 东凌经济管理学院  
王柏琳 北京科技大学 东凌经济管理学院  
中文摘要
      本文从无缝钢管生产实际中提取并定义了周期性机器检修环境下的钢管热轧批量计划问题, 基于无缝钢 管生产的特殊性, 将该问题抽象为一类考虑机器检修和机器调整时间的单机调度问题, 并建立了以最小化机器闲置 和机器调整时间为目标的数学模型. 针对批量间的机器调整时间取决于钢管规格的变化这一特性, 提出了最小调 整时间排序规则, 证明了该规则在不考虑检修计划时具有最优性. 进而, 以此为基础建立了循环求解框架, 并设计 了两阶段启发式算法. 基于实际生产数据设计了多种问题规模的实验, 验证了算法的有效性, 并从实际应用角度对 结果进行了分析.
英文摘要
      In this paper, a hot-rolled batch planning problem for seamless steel tube was extracted and defined from actual production under the condition of periodic machine maintenance. Due to the particularity of the seamless steel tube production, the problem was abstracted into a single machine scheduling problem with machine maintenance and setup times, and its mathematical model was established to minimize the total idle time and total setup time. Based on the feature that the setup time between batches depends on the specifications of steel tubes, a minimum rolling mill setup time rule was proposed, and it was proved to be optimal when the maintenance plan is not considered. In addition, a solving strategy and a cyclic solving architecture were furthermore established based on it, and a two-stage heuristic algorithm was designed. Finally, based on actual production data, experiments of different scale were carried out to evaluate the performance of the algorithm. The experimental results show that the algorithm can get a near optimal solution in a short time. Moreover, these results had been accordingly analyzed from practical point of views as well.