【作者】 许文砚；

【导师】 叶文华；

【作者基本信息】 南京航空航天大学 ， 机械电子工程， 2005， 硕士

【摘要】 随着市场竞争的日益剧烈,大多数制造企业的生产方式正向着多品种小批量方向发展,因此面向单件小批量生产方式的 Job Shop 调度问题已成为一个具有代表性的生产调度问题。本文在传统 Job Shop 车间作业调度的基础上,放宽了资源约束条件,针对这种比较复杂的组合优化问题,采用基于规则的调度算法,选择工件相对剩余加工时间最大和在设备上等待时间最短两条优先规则,分别解决了多个零件工序竞争一台设备时的工件排序问题和一个零件工序可在多台设备上加工时的设备选择问题,经过实例验证,获得了较为理想的结果。同时,本文分析了数据挖掘技术在车间调度问题中的应用前景,并提出了利用 ID3 决策树方法提取组合规则的运算思想,从调度历史数据中提取出隐含的排序规则来解决加工冲突问题,通过具体实例阐述此算法的详细运算步骤并验证了其可行性。最后,本文阐述了基于单件小批量模式的车间生产调度系统的设计思想,利用具体建模工具对系统进行分析和建模,详细介绍了该系统的功能模块与关键技术,并开发了较为实用的生产调度原型系统。更多还原

【Abstract】 The production mode of most manufacturing enterprises is turning to small batchalong with the intense competition of the market, so the job shop scheduling problemhas become a typical scheduling problem. The problem investigated in this paper is an extension of the classical job shopscheduling problem that relaxes the restriction of resource. The paper firstlyintroduces an algorithm based on priority to solve this complicated combiningoptimization problem. The rule of maximal comparative remaining operating time isused to choose the part when operations of different parts compete one machine. Therule of least waiting time of the part on a machine is used to choose the machinewhen an operation of a part has more than one usable machine. This algorithm isproven to be effective by examples. Secondly, the paper analyses the application of data mining in the schedulingproblem. It puts forward the method of finding combinatorial rules by decision treebased on ID3 algorithm. The hidden rules of sequencing can be mined form historicaldata to solve the problem of operations of different parts competing one machine. Bymeans of practical examples, the detail steps of this algorithm are expatiated and itsfeasibility is proven. Finally, the paper discusses the design scheme of production scheduling systemin the small batch workshop. The system is analyzed and designed by modeling tools.The function modules and the realizing technologies are introduced in detail. Aprototype system is developed.更多还原