• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Japan/USA Symposium on Flexible Automation >REACTIVE JOB-SHOP SCHEDULING BY USE OF AN ARTIFICIAL NEURAL NETWORK WITH FUZZY TEACHING
【24h】

REACTIVE JOB-SHOP SCHEDULING BY USE OF AN ARTIFICIAL NEURAL NETWORK WITH FUZZY TEACHING

机译：通过使用具有模糊教学的人工神经网络的反应性工作店调度

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

This paper proposes a reactive job-shop scheduling method using a back propagation neural network with fuzzy teaching. A production state is defined by the combination of three fuzzy numbers. Each fuzzy number represents a different estimation of the job-shop's state. Teaching data is a set of state and dispatching rule pairs. When some state is given, a genetic algorithm finds out the optimum or quasi-optimum schedule. At the same time, dispatching rules make their schedules, respectively. By comparing the make-span time obtained by the genetic algorithm with ones obtained by dispatching rules, the dispatching rule making the closest schedule to one obtained by the genetic algorithm is chosen for the given state as a teaching datum. Once the neural network learns a set of teaching data, it determines an adequate dispatching rule in responding on any production state. Numerical simulation results verify that the performance of the proposed method is similar to that of the genetic algorithm and better than that of single dispatching rules with regard to the maximum and mean flow time.
机译：本文提出了一种利用具有模糊教学的后传播神经网络的反应性工作店调度方法。生产状态由三个模糊数的组合定义。每个模糊数表示作业商店状态的不同估计。教学数据是一组状态和调度规则对。当给出某些状态时，遗传算法发现了最佳或准优选的时间表。同时，分派规则分别进行他们的时间表。通过比较通过调度规则获得的遗传算法获得的制造跨度时间，为给定状态选择使得最接近的时间表成为由遗传算法获得的调度规则作为教学基准。一旦神经网络了解一组教学数据，它就决定了响应任何生产状态的充分调度规则。数值模拟结果验证所提出的方法的性能与遗传算法的性能类似，并且比在最大和平均流动时间方面的单一调度规则更好。

著录项

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