【作者】 郭宏；

【导师】 庞思勤；

【作者基本信息】 北京理工大学 ， 机械制造及其自动化专业， 2015， 博士

【摘要】 刀具作为制造企业完成切削加工的重要工艺装备,虽然只占到了加工成本的4%,但合理的刀具使用可以使生产成本降低25%左右。自1964年第一个切削数据库建立以来,人们就在不断积累、总结刀具使用过程中的宝贵数据,但是直到现在这些分属于刀具厂商、制造企业和科研院所的大量切削工艺数据及刀具实体数据还未形成统一的数据共享使用模式,而这已成为构建“服务型”制造系统必须解决的关键问题之一。本文针对刀具使用中数据共享不畅的现象,提出面向制造的刀具资源数据服务研究。在详细分析了刀具相关数据的来源、分类、结构、内容和使用形式的基础上,提出面向服务架构的刀具资源数据服务体系,建立由刀具制造者--刀具厂商,刀具使用者--制造企业,刀具使用技术支撑者--科研院所构成的为制造企业提供刀具资源数据服务的系统。在分析了刀具资源数据在制造过程中的使用现状的基础上,研究了面向制造的刀具资源数据服务体系的建立过程及若干关键问题。对数据服务体系按照层次划分和服务生命周期管理阶段划分,并分别进行描述:提出采用刀具资源数据服务提供层,刀具资源数据服务管理平台和刀具资源数据需求层的三层架构的刀具资源数据服务模型方案,并对每层的功能进行详细分析说明;按照服务生命周期管理定义,建立包括刀具资源数据服务的需求分析、服务的定义及分析、服务匹配和监控、服务发布、服务评估在内的刀具资源数据服务管理周期并就各阶段完成的内容进行具体分析。主要工作和研究成果如下:(1)刀具资源数据的服务化迁移过程研究针对现有刀具资源数据与相应映射服务的多重性与模糊性,分别建立刀具资源数据及数据服务描述立方体,采用在立方体模型内不同维度间的信息迁移及模型间共通维度的数据函数迁移的方法,完成刀具资源数据--服务映射,建立包括刀具资源数据、服务、运算规则和迁移规则的迁移过程形式化描述模型,为实现标准化、规范化、一致化数据服务描述奠定了基础。(2)刀具资源数据服务匹配策略研究为提高数据服务的有效性,以刀具资源本体模型为基础,建立由三种匹配原则和考虑个人喜好的排序规则构成的服务匹配机制。首先依据加工方式不同对刀具资源进行分类,并以此为基础完成了刀具资源本体模型的构建。以刀具资源本体模型为匹配依据,完成包括文本语义距离、语义重合度和语义深度的综合语义文本匹配模型的设计;以刀具资源服务功能为匹配依据,通过主观评判法建立的模糊权重匹配表,完成服务功能匹配模型的设计;依据刀具资源数据属性特性,完成刀具资源数据属性匹配模型的设计。最后依据考虑个人喜好的Borda投票排序法,对三种单原则匹配下的排序结果进行综合,获得最优排序。(3)多类型参数的刀具切削性能预报方法研究针对多类型参数对刀具切削性能影响没有普适经验公式的情况,提出建立动态刀具切削性能预报的方法。结合遗传规划的树形结构直观性与符号回归的多样性,提出建立包括刀具设计参数,制造参数和切削(使用)参数的多类型参数的刀具切削性能预报方法。该方法已应用在高速钢丝锥寿命预报中,并对预报结果与线性预报结果进行了比较,验证了方法的有效性。(4)刀具资源数据服务评价方法研究按照综合设计模式,采用分析模式构建由目标层,子目标层和指标层组成的三层刀具资源数据服务综合评价体系模型。评价过程中为了体现用户需求的差异性,建立客观服务内容评价和主观服务效果评价相结合的组态式定制服务评价方法,为后续刀具资源数据服务的选择提供依据。(5)开发基于面向服务架构的刀具资源数据服务平台基于Internet和浏览器技术的B/S结构,采用.NET技术架构,开发了基于面向服务架构的刀具资源数据服务平台。对平台的层次架构及数据服务功能实现技术进行了说明,并通过实例对刀具资源数据服务过程中的相关理论和服务方法进行验证,验证各章理论方法的可行性和实用性。更多还原

【Abstract】 The cutting tool is the important equipment to finish cutting in manufacturing. Although its cost only accounts for 4% of process spending, the appropriate choice of it may reduce the production cost about 25%. From the establishment of the first cutting tool database in 1964, people began to collect the corresponding data in the course of using and try to gather experience to use it. But until nowadays, the technical data are still separate in tool manufacturers, users and researcher institutes, and there is no unified data sharing models, which has been a key problem in constructing the service-oriented manufacture system.According to the obstructed data-sharing phenomenon in the course of using, this paper studies on tools resource data service(TRDS) in manufacturing. After analyzing the source, classification, structure, content and use-form of tools data in detail, it establishes the SOAoriented data service system and the data service to manufacturing enterprises composed of tool makers, tool manufacturers, tool users and manufacturing enterprises and tool researcher— the academic institutions.After analyzing the data utilization condition in manufacturing, the paper focuses on the establishment of TRDS system. The data service system according to the hierarchy and the periods of service lifecycle management respectively. The corresponding 3-layer-architecture tool resource data service model scheme is proposed and explained, which consists of the providers, the management platform and the demanders of the tool resource data service. The service management lifestyle composed of requirement analysis, definition analysis, service match and monitor, service announcing and service evaluation is established and the main task within each period is stated specifically.The paper is organized by follows:(1) Tool resource data servitization transfer modelIn order to avoid the multimap and fuzzy between tool resource data and corresponding data service, tools resource data and service description cubes are modeled respectly and the mapping is established by the information transfer among dimensions in one cube and the universal description dimension between cubes. Besides data and data service cubes, together with the rules for serivitization process including operating and transferring, the transfer description model is established. The model attributes to describe the data service strictly and uniformly.(2) Match strategy for TRDSBased on tool resource ontology model, service match mechanism consists of three match steps and an order rule, including text match, function match, property match and the order rule considering personal preference. Tool resources are classified according to the machining mode, which is the basis for tool resource ontology model. In accordance with the ontology model, the match of text distance, coincidence and depth are completed. The fuzzy weighted matching table based on subjective evaluation method is utilized in function match. Difference of data property plays an important role in property match. Borda, considering personal preference, is applied to rank the candidate after they are ordered with each single matching rule.(3) Tool cutting performance prediction with multi-type parameters.Owing to no experience formula for multi-type parameters, including tool manufacturing parameter, heat-treatment parameter and usage parameter, a dynamic tool cutting performance predication method is put forward. Visuality of tree structure of genetic programming and multiformity of symbolic regression are combined and has been application in high steel tap life prediction. By comparing the statistics parameters on tap life achieved from the novel settlement and liner one, it demonstrates the effectiveness of the proposed method.(4) Evaluation method of TRDSThe evaluation system comprises of object layer, subobject layer and index layer. According to individual requiement, the demand user limits index and the system accomplishes the subjective evalue of service effect and the objective evalue of service content independently or together. It provides evidence for tool resource data service option.(5) Implement of TRDS platform based on SOAThe relevant software of TRDS is designed and implemented based on Browser/Server(B/S) and.NET. The platform layer frame and application technology are stated. An example illustrates how the system works and proves the practicality and feasibility of theoretic approach presented above.更多还原