• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文学位 >Heuristic approaches to minimize the makespan and tardiness: Scheduling jobs with job splitting property on uniform parallel machines with different capacities.
【24h】

Heuristic approaches to minimize the makespan and tardiness: Scheduling jobs with job splitting property on uniform parallel machines with different capacities.

机译：启发式方法，最大程度地减少了工期和拖延时间：在具有不同容量的统一并行计算机上调度具有作业拆分属性的作业。

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

This thesis aims to develop heuristic approaches to Schedule Jobs with job splitting property on Uniform parallel machines with different sizes. Job Splitting property has been used to develop heuristics for efficient scheduling in the research community. However, the setting where the machines of different sizes with capability of processing any type of jobs have not been researched. This was the motivation behind approaching this type of problems. Three heuristics are developed based on proportional splitting. After the application of dispatching rules of earliest due dates and shortest processing time, the heuristics are then applied to twelve simulated problem instances to evaluate their performance. The problem instances consider different job sizes and number of jobs, and different machine sizes and number of machines. The heuristics that have better performance are combined with a simple genetic algorithm (GA) to further improve their solution qualities. The parameters of GA are tuned using design of experiments with 2 levels of 4 factors (number of iterations, population size, probability of crossover, and probability of mutation). The experimental results of the heuristics are statistically compared at 95% confidence level. The heuristic 2, which is an extension of heuristic 1, in which the largest machine is considered as critical and made to potentially run one more time to process a part of the unscheduled section of the job. Using heuristic 2 along with the GA provides efficient and consistent results. The application of heuristics with different combination of dispatching rules can also be explored in future.
机译：本文旨在开发启发式方法来调度具有不同大小的均匀并行机上的作业拆分属性的作业。作业拆分属性已用于开发启发式方法，以在研究社区中进行有效的调度。但是，尚未研究具有处理任何类型作业能力的不同尺寸机器的设置。这是解决此类问题的动机。基于比例分裂，开发了三种启发式方法。在应用最早的到期日期和最短的处理时间的调度规则之后，该启发式方法随后被应用于十二个模拟问题实例以评估其性能。问题实例考虑到不同的作业大小和作业数量，以及不同的机器大小和机器数量。性能更好的启发式算法与简单的遗传算法（GA）相结合，可以进一步提高其求解质量。 GA的参数使用2个级别的4个因子（迭代数，种群大小，交叉概率和突变概率）的实验设计进行了调整。启发式的实验结果在95％的置信水平上进行了统计比较。启发式2是启发式1的扩展，其中最大的机器被认为是至关重要的，并且有可能再运行一次以处理作业的计划外部分。将启发式2与GA结合使用可提供有效且一致的结果。将来还可以探索启发式方法与调度规则的不同组合。

著录项

  • 作者

    Shetty, Punit N.;

  • 作者单位

    State University of New York at Binghamton.;

  • 授予单位 State University of New York at Binghamton.;
  • 学科 Operations research.;Industrial engineering.
  • 学位 M.S.
  • 年度 2015
  • 页码 69 p.
  • 总页数 69
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类 水产、渔业;
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号