【作者】 韦韫；

【导师】 李东波；

【作者基本信息】 南京理工大学 ， 机械制造及其自动化， 2011， 博士

【摘要】 网络化协同制造是适应网络经济和知识经济的先进制造模式,它强调企业间的协作和资源共享,从而实现降低生产成本和提高资源利用率的目标。网络化协同制造资源的重组优化配置是实现网络化协同制造的关键技术,也是现代集成制造领域研究的热点问题之一。在服务经济和循环经济的全球化发展驱动下,制造业逐渐呈现出采用服务化满足客户服务需求的趋势,出现了面向服务的制造。由于面向服务技术能够很好的满足面向服务制造的需求,将面向服务技术引入网络化协同制造,构建支持全球资源共享和协作的面向服务的网络化协同制造系统,具有重要的理论价值和现实意义。本文主要研究了基于面向服务架构的网络化协同制造资源重组优化系统体系结构以及关键技术,主要包括如下几个方面：(1)研究了网络化协同制造资源重组优化配置系统体系结构。在对面向服务的网络化协同制造资源重组优化配置系统需求分析的基础上,基于面向服务架构和多Agent(代理)技术提出了面向服务的网络化协同制造资源重组优化配置系统体系结构。在给出了该体系结构各层内容后,设计了系统的功能模型和IDEFO模型,最后分析了系统实现的关键技术。(2)研究了基于元模型的网络化协同制造资源建模。在分析了网络化协同制造资源建模的需求的基础上,针对网络化协同制造资源的特点对资源进行了分类并分析了资源建模对面向服务平台的需求。在此基础上提出了协同制造资源元模型,包括基础数据元模型、业务过程元模型和应用扩展元模型三方面,并在元模型中引入了与情境相关的元类,实现了元数据级制造资源建模。然后,分析了网络化协同制造资源模型的建立过程及模型的集成和扩展。最后,基于XML对制造资源进行了描述。(3)研究了面向用户需求的网络化协同制造资源发现与检索。提出了一种基于灰色模糊聚类的多层次网络化协同制造资源发现与检索方法：首先提出了面向用户需求的基于灰色关联度的资源模糊聚类方法——根据用户需求实现了基于灰色关联度的模糊聚类,并建立了面向用户需求的资源聚类框架。提出了以资源簇内聚合度、资源簇间分离度和用户满意度线性组合为资源聚类优化目标的资源聚类质量评价数学模型,并运用GA进行了优化,以应用实例证明了该聚类方法的有效性。提出了一种基于资源类型的多层次分级检索方法,建立了面向用户需求的分布式多层次资源发现框架,框架分为自组织域层、聚类层和扩散层。最后,建立了相应的检索流程。(4)研究了面向服务的网络化协同制造资源多目标重组优化调度。针对网络化协同制造资源重组优化调度中存在的问题,综合考虑影响优化调度的四个主要因素,即生产作业时间、生产作业成本、生产加工质量和资源服务质量,提出了网络化协同制造资源多目标重组优化调度的数学模型。同时,提出了一种基于Pareto多目标免疫遗传算法的网络化协同制造资源重组优化调度方法,该算法综合运用了小生境技术、群体排序技术和精英保留策略,并对遗传算子进行改进,自适应的调整交叉和变异算子,结合免疫算法的免疫选择淘汰了相似个体,保证了种群多样性,避免了早熟现象的发生。免疫记忆对近似最优解进行动态邻域搜索,提高了算法的局部搜索能力。(5)研究了网络化协同制造资源重组评价体系。根据网络化协同制造资源重组优化的特点和评价指标的选取原则,提出了较为完备的网络化协同制造资源重组优化的评价体系模型。在分析了各指标含义的基础上提出了一种基于灰色理论和模糊集的网络化协同制造资源重组优化评价方法实现了对各指标的评价。采用灰色关联系数作为网络化协同制造资源评价指标的隶属度,它不仅考虑了各指标自身的影响,也同时考虑了相关指标对其优属程度的影响,因而体现了方案的整体性。(6)建立了基于面向服务架构的网络化协同制造资源重组优化配置系统集成框架,并设计了系统的Agent结构和各Agent之间的通信机制。在上述关键技术研究的基础上,进行了系统原型的开发,该系统已在某航天企业中得到初步应用,效果良好。更多还原

【Abstract】 As a kind of advanced manufacturing paradigm in network economy and knowledge economy environment, collaborative manucturing emphasizes inter-enterprise collaboration and resource sharinig in the whole society.The goal of collaborative manufacturing is to reduce the production cost and improve the resource efficienvy. Networked cooperative manufacturing resource reconfiguration and optimization is the key technology to realize distributed collaborative manufacturing. It has become one of the hot topics in the study field of modern integrated manufacturing. Under the drive of the global development of service economic and circle economic, the manufacturing puts up the trend of serving the customer demand, and the oriented-service manufacturing appears. Due to the satisfaction of requirement of service-oriented manufacturing, it is of great the theoretical value and realistic significance to construct service-oriented networked cooperative manufacturing system to support global resource share and collaboration with service-oriented technology introduced into networked cooperative manufacturing field. This dissertation mainly discusses the architecture and key technologies of networked cooperative manufacturing resource reconfiguration optimization system. The main research contents in this dissertation are as follows:Firstly, the overall framework for NCMROR is studied. After the system requirement is anaylzed for the NCMROR, the system architecture is presented based on the service-oriented architecture and multi-agent technology. The main contents of the system structure are analyzed, and the function model of system is designed. The key of system realization are discussed and thus the research direction of the dissertation is determined.Secondly, the modeling technology of networked cooperative manufacturing resources based on the meta-model is studied. After the analysis of networked cooperative manufacturing resources modeling, the resources are classified and the requirement is analyzed for the modeling of cooperative manufacturing resources considering the characteristics of cooperative manufacturing resources. The ooperative manufacturing resources meta-model which was composed of three sub-meta-models including the basic data meta-model, the operation process meta-model and application meta-model was presented. Then, analysis of cooperative manufacturing resource information model proved its ability of resource model integration and extension. At last, the resource was described by eXtensible Markup Language.Thirdly, the customer-demand-oriented discovery and search of networked cooperative manufacturing resources(NCMR) are studied. A hierarchical NCMR search algorithm is proposed based on the gray and fuzzy clustering:at fist, On the basis of gray related criterion and customer demand, a fuzzy clustering algorithm is presented——the fuzzy clustering based on the gray related criterion is realized and the framework of the NCMR fuzzy clustering based on the customer demands. After the clustering algorithm was proposed, an optimization model of NCMR clustering evaluated by internal clustering degree、external detached degree and the satisfying degree of customers is proposed and genetic algorithm is applied to optimize the clustering results. Furthermore, an application instance proves the validity of the algorithm. After NCMR being clustered, a hierarchical、distributed search and customer-demand-oriented framework of NCMR is structured including self-organizing laye、clustering layer and diffused layer. Finally, the corresponding flow of is established.Fourthly, the multi-objective reconfiguration optimization scheduling of NCMR service-oriented. Considering the problem exists in the reconfiguration optimization scheduling of NCMR, a the multi-objective reconfiguration optimization scheduling of NCMR mathematical model is proposed, which includes four main factors. Such as production working time, production working cost, production process quality and resources service quality are considered. A scheduling method for the multi-objective reconfiguration optimization scheduling of NCMR is presented based on Pareto multi-objective immunity genetic algorithm. In order to ensure the groups variety, prevent the premature convergence problem, some key technologies such as population ranking technique, niche technique, Pareto solution set filter were applied, the algorithm improved the genetic operators and proposed self-adaptive crossover and mutation operators, and immune selection could eliminate semblable individuals. Immune memory was adopted to dynamic search near each approximate optimal solution, so it accelerated search and improved quality of solution. Finally the result of simulation indicates validity of algorithm.Fifthly, the evaluating system for NCMR reconfiguration is studied. On the basis of the analysing of the characteristics of the NCMR reconfiguration and criterion of the evaluation index, a comprehensive evaluation model for NCMR reconfiguration is proposed. After the analyzing of the meaning of every index, a evaluation method of NCMR reconfiguration optimization based on the gray theory and fuzzy muster is proposed and the evaluation for every index is realized. Finally, an instance for resources reconfiguration project optimal selection of some enterprise proves the validity of the evaluation system.Sixthly, the integrated framework of the NCMR reconfiguration optimization system based on the architecture service-oriented is presented. The Agent structure of system and the communication mechanism between every Agent are designed.Finally, on the basis of above research, the system prototype of NCMR reconfiguration optimization is developed. The system is applied in the engineering, and the result shows the effectiveness of the system.更多还原