作业车间实时调度若干关键问题研究

发布时间：2018-06-28 16:11

本文选题：车间作业调度 + 实时调度　；参考：《重庆大学》2013年博士论文

【摘要】：车间生产调度系统是先进制造系统中最为重要的内容之一，它对资源优化利用，提高工作效率，节省成本都有非常重要的作用。随着经济的快速发展，制造企业之间的竞争越来越激烈，制造企业也越来越注重通过高效的车间生产调度系统来提高企业的竞争力。经典作业车间调度问题的研究虽然取得了丰富的理论成果，但由于所建立的模型与生产实际状况相差甚远，因此很难用于指导生产实践。近年来对作业车间实时调度问题的研究逐渐成为制造系统研究领域的一个热点，该问题的研究具有重要的理论意义和工程实用价值。本文对作业车间实时调度的某些关键问题进行了深入的研究，研究工作的主要内容如下。 ①当前制约作业车间实时调度技术应用于车间现场的瓶颈是，缺乏一个合适的能够对作业车间实时调度系统及其优化算法进行有效仿真和验证的平台。针对这个问题，论文中提出了基于RFID的实时调度物理仿真平台的框架结构，期望通过该平台为作业车间实时调度问题的理论研究提供一个仿真和验证的环境。论文建立了平台的总体框架，对平台中涉及到的各个组成部分，包括：RFID卡组件、RFID卡读写器组件、RFID卡盒组件、RFID卡传送装置组件、通信网络的架构、数据信息格式等内容进行了讨论。 ②提高生产调度系统的柔性是车间实时调度系统建模必须考虑的一个问题。针对这个问题，在对工序树、工件叶子集合、空闲设备集合、作业进程表等概念进行详细分析的基础上，论文提出了一种基于工序树的工艺规划和调度系统集成的方法。其中用于实验和验证的程序，经过完善和扩充可以运行在基于RFID的实时调度系统仿真平台上，同时该程序中的调度子模块可以很方便地嵌入各种调度优化算法。 ③寻找效果更佳的调度优化算法是车间实时调度系统理论和方法研究中的一个基本问题。对于这个问题，论文中提出了两种作业车间实时调度优化算法，具体的内容如下： 1）提出了一种基于组合规则的实时调度方法。组合规则调度的难点在于如何对规则进行组合才能最好地达到调度目标的要求。本文的解决办法是，，先建立工件选择规则库和设备选择规则库，然后根据调度目标要求用AHP法对规则库中的规则进行排序，最后再依据规则库中的排序结果进行规则的组合。实验的结果验证了该方法可以起到作业调度优化的作用。 2）提出了一种基于自适应规则的实时调度方法。由于调度规则固定的调度方法在车间实时调度中不能适应环境的变换，该方法在实时调度中的应用效果不好。因此，论文中用强化学习的方法对调度规则进行训练来实现自适应规则调度。在自适应规则调度中，环境被映射成一定的模式或者状态，调度算法根据环境的不同，智能地选择该环境状态下最合适的调度规则。实验的结果验证了该方法可以起到一定的调度优化作用。
[Abstract]:The workshop production scheduling system is one of the most important contents in the advanced manufacturing system. It plays a very important role in optimizing the utilization of resources, improving the efficiency and saving the cost. With the rapid development of the economy, the competition among the manufacturing enterprises is becoming more and more intense, and the manufacturing enterprises pay more and more attention to the efficient workshop production scheduling system. The research of the classic job shop scheduling problem has made a lot of theoretical results, but it is very difficult to guide the production practice because the model is quite different from the actual production situation.
In recent years, the research on the real-time scheduling problem of the job shop has gradually become a hot spot in the research field of the manufacturing system. The research of this problem has important theoretical significance and practical value of engineering. This paper has carried out a thorough study on some key problems of the real-time scheduling of the job shop, and the main contents of the research work are as follows.
(1) the bottleneck that restricts the application of real-time scheduling technology to the workshop site is the lack of a suitable platform for the effective simulation and verification of the job shop real-time scheduling system and its optimization algorithm. In this paper, a framework of the real time scheduling physical simulation platform based on RFID is proposed in this paper. The platform provides a simulation and verification environment for the theoretical research of the job shop real-time scheduling problem. The paper establishes the overall framework of the platform, which includes all the components involved in the platform, including RFID card components, RFID card reader components, RFID card box components, RFID card transport components, communication network architecture, data letter. The content of the interest format is discussed.
To improve the flexibility of the production scheduling system is a problem that must be considered in the modeling of the real-time scheduling system of the workshop. On the basis of the detailed analysis of the concept of process tree, the set of workpiece leaves, the set of idle equipment, the operation process table and so on, a process planning and scheduling system integration based on process tree is proposed in this paper. The program, which is used for experiment and verification, can be perfected and expanded on the RFID based real-time scheduling system simulation platform. At the same time, the scheduling sub module in the program can easily embed various scheduling optimization algorithms.
To find a better scheduling optimization algorithm is a basic problem in the study of the theory and method of the workshop real-time scheduling system. For this problem, two kinds of job shop real-time scheduling optimization algorithms are proposed in this paper, and the specific contents are as follows:
1) a real-time scheduling method based on combination rules is proposed. The difficulty of combination rule scheduling is how to combine rules to best achieve the requirements of the scheduling goal. The solution of this paper is to establish the selection rule base of the workpiece and the selection of the rule base of the equipment first, and then according to the requirement of the scheduling target, the AHP method is used for the rule base. The rules are sorted, and then the rules are combined according to the ranking results in the rule base. The experimental results show that the method can play a role in the optimization of job scheduling.
2) a real-time scheduling method based on adaptive rules is proposed. Because the scheduling rules can not adapt to the transformation of the environment in the real-time scheduling of the workshops, the application effect of this method is not good in real-time scheduling. Therefore, the method of reinforcement learning is used to train the scheduling rules to realize the adaptive rule scheduling. In the adaptive rule scheduling, the environment is mapped to a certain pattern or state. The scheduling algorithm selects the most appropriate scheduling rules in the environment according to the different environment. The experimental results verify that this method can play a certain role in scheduling optimization.
【学位授予单位】：重庆大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：TH186

【相似文献】