【作者】 王东强；

【导师】 鄢萍；

【作者基本信息】 重庆大学 ， 机械制造及其自动化， 2010， 博士

【摘要】 现代作业车间是一种先进制造系统,其生产过程是一个多设备多任务并存的过程,具有生产品种多、个性化程度高,变化快,批量较小等特点。作业车间在运行过程中要实现对生产计划的快速响应和生产任务的优化调度,必须随时动态掌握车间制造系统中在线制造的工件即在制品的动态情况,设备的运行状态及利用情况等一系列信息状况,因此必须从系统的角度研究这些生产进度信息的内在关联、动静态特性、提取及分析方法。论文从现代作业车间制造系统的优化运行过程角度,以其制造过程“快速、优质、高效、低成本、低耗”的综合需求为目标,结合国家自然科学基金课题“现代作业车间制造系统运行状态信息的系统特性和采集新方法研究”,(项目编号:50775228)。对车间复杂生产进度信息的网络化特性、提取方法及应用模型等内容进行了研究。主要研究内容有:通过分析作业车间制造系统生产进度信息的特点,采用复杂网络的思想首次给出了作业车间复杂进度对象的网络化描述,建立了适应多订单任务和多设备的制造系统复杂生产进度网络的分层模型和信息内容,并采用节点度值,聚类系数及冗余度等相关网络化特征参数进行描述。为了解决如何反映当前复杂进度信息网格中单个节点进度状态的问题,论文首次提出了一种基于Markov链和信息熵理论的作业车间信息节点进度特征分析算法。将节点多元进度信息特征划归为三类,即强相关特征、弱相关特征和无关特征,同时利用模糊层次分析法深入分析根据信息节点最优运行进度信息集合中各组成要素之间的直接关系和间接关系,以及各进度子类中各要素对整个系统的影响程度和每个要素在整个系统结构中的重要程度。通过设置特征阀值的方法利用相关性水平过滤剔除特征集合中的无关和冗余特征,简化为作业车间信息节点指定时刻或阶段信息节点关键路径相关的最优进度信息集合。利用作业车间进度信息网络化描述模型建立了作业车间复杂进度信息关系网及其信息提取方法。定义了作业车间复杂进度信息关系网的生成机理、演化机理及演化算法,将复杂生产进度信息提取问题转换成复杂网络上的节点遍历问题,这为研究作业车间实际复杂生产进度问题提供了新视角和新方法。在上述理论研究成果的基础上,开发了一种作业车间制造系统生产进度信息智能提取及分析的支撑系统原型,用于辅助车间生产管理人员在安排调度方案或车间监控时,对加工过程中的动静态加工特性、资源消耗、人员流动等相关因素进行综合的信息提取及分析。并对其基本的运行流程、功能模块设计以及软件架构设计等进行了分析,该系统具有动态在线显示和在线控制的功能,实现了作业车间设备加工进度数据、资源进度信息、生产物流等生产运行状态信息与上层应用系统之间的无缝交互与集成,解决了多元信息的融合和分析、多信息的集成、大信息量交互、与远程需求部门的实时无障碍交互等系列难题。最后结合某制造企业的作业加工车间对该系统的应用情况进行了介绍。更多还原

【Abstract】 Modern job shop, whose production process is the coexistence of multi-device and multi-task process, is an advanced manufacturing system with many features such as various kinds of production, a high degree of personalization, rapid change, small volume and so on. During the operation of Job shop, rapid response to production plan and optimal scheduling of production tasks must be achieved, and a series of information status must be ready to grasp, such as the dynamic situation of online manufacturing workpiece, the condition and utilization of equipments, etc. Thus, the internal relationship, dynamic characteristics, extraction and analysis methods of the information in the shop manufacturing systems should be researched from the system point of view. With the comprehensive needs of "Fast, high quality, high efficiency, low cost, low power consumption" as the goal, and in combination with the National Natural Science Fund Project, Research on System Characteristics and New Collecting Methods of the Operating Status Information of Modern Job Shop Manufacturing System(Project Number:50775228), extraction methods and application models of the plant operating status information are studied in the paper from the view of optimal operation of the modern job shop manufacturing system. Main research contents and results are listed bellow:The production process information of the job shop manufacturing system is analyzed. A networked description of the job-shop complex progress object is given with the idea of complex network for the first time. A hierarchical model of complex production progress network is established to adapt to the product tree of ordering task and the dynamic changes in the implementation process. The node granularity of the complex progress information network is defined. At the same time, three kinds of local feature indicators of the complex progress information networks and analyzed, including node degree value, clustering coefficient and redundancy.In order to solve the problem of how to reflect the running state of the information node in job shop, this paper provided a kind of analysis algorithm on progress of information node in job shop based on Markov Chain and Information Entropy theory for the first time. The multiplex progress information characteristics are divided into three groups, namely strongly correlated characteristics, weakly correlated characteristics and uncorrelated characteristics, and the concept of Entropy in information theory is refereed to here to simplify the aggregation of characteristics to aggregation of optimal progress information of information nodes in job shop, through setting threshold level of characteristic to filter the correlation extent of characteristic to eliminate uncorrelated and redundant elements in the original aggregation of characteristics. At the same time,hierarchical fuzzy analysis method is used to deeply analyze the direct an indirect relations between elements in the aggregation of optimal operational schedule information, extent of affection from each element in all progress subclasses to the whole system, and importance of which in the system architecture.This paper designed the detailed acquiring steps and some related rules, and then, the complex network itself and the acquiring method are validated in this paper through simulation experiment on the progress information relation network of job shop. One networking method of acquiring complex process objects in job workshop is researched in this paper, which converted the problem of acquiring complex progress information to a problem of traversing nodes in complex network, and that provided a new perspective for researching complex production process problem.Based on the above research results, a smart prototyping support system for acquiring and analyzing multiplex information in job shop has been developed, which can be used to assist workshop managers to comprehensively acquire and analyze the dynamic and static characteristics, resource consume, manpower flow, and some other related factors while they are making scheduled plan and monitoring production process. Additionally, this paper elaborated the architecture and function modules design of this prototyping system, and analyzed its basic operational process. The system provide with the function of dynamically display on line and control on line, and realized the seamless interaction and integration, and resolved a series of problems, such as merge and analysis of multiplex information, integration of multi-message, huge information interaction, real-time and barrier-free interaction with remote requesters. At the end of this paper, introduction about the use of the prototyping system in job workshop of certain manufacturing company is given.更多还原