SIA OpenIR  > 工业控制网络与系统研究室
基于本体的自动化生产线建模方法及其应用
Alternative TitleAn ontology-based modeling approach of automated production line and its application
施昭
Department工业控制网络与系统研究室
Thesis Advisor于海斌
Keyword本体 自动化生产线 服务编排 语义标注 情景响应和处理
Pages102页
Degree Discipline机械电子工程
Degree Name博士
2018-08-30
Degree Grantor中国科学院沈阳自动化研究所
Place of Conferral沈阳
Abstract

随着互联网的快速发展和信息通信技术的不断成熟,模块化的思想、开放性和跨平台性的特点,为解决生产制造系统的重配问题和提高生产制造系统模型的灵活性和适应性提供了一种可行的技术支持。当前,在工业制造领域中,一些研究致力于应用面向服务或Agent技术的方法,将生产设备作为类似于网页这样离散的实体进行描述,通过Web服务或Agent对生产设备的模块化抽象封装和本体的描述,克服了传统PLC设计中可重配空间小的局限性,促进了异构生产设备的互联互通,基于BPEL流程实现了资源的自动发现、调用和编排,提高了生产制造系统的灵活性和适应性,降低了系统重配的代价。然而,BPEL是一种业务过程执行语言,是以整体流程的方式表达某个生产对象的制造过程。这样的表达方式,结构性比较强,灵活性和扩展性有限。当产品的生产制造过程发生变化时,整个BPEL流程都需要进行重新的人工设计,而BPEL设计、部署、编译的过程非常复杂,这就给系统的重配带来了非常大的代价。另外,基于BPEL的服务匹配和编排,是一种类似于互联网上网页查找和匹配的过程,不但时耗比较长,而且会涉及查全率和查准率的问题,对此这些研究方法并没有进行深入的研究。 因此,为了让有序化结构特征在表达上具有更好的灵活性和扩展性,能够基于产品特征灵活的对产品的制造工艺流程进行编排,以批量化的生产能力完成个性化产品的定制化生产,同时促进生产制造系统对实时数据的理解和分析,能够基于情景对产品的生产过程做出动态的响应和调整,本文在已有研究的基础上,提出了一种基于工艺特征的生产线本体模型架构和自动化生产线本体管理及服务编排系统框架,构建了工艺特征本体和数据属性本体,并提出了基于本体模型的服务匹配编排方法和情景响应处理方法。通过在本体模型中对制造工艺知识、设备次序关系和数据属性特征的描述,提高了生产制造有序特征和数据属性特征在表达方式上的灵活性和扩展性,以确定性和非歧义的表达方式促进了机器对有序特征和信息含义的准确性理解,为生产制造系统基于情景的智能化响应和处理提供技术支撑,在生产制造系统从层次化结构向扁平化结构转变的基础上,实现了继续向有序化结构转变的过程。 首先,本文提出了一种基于工艺特征的生产线本体模型架构,通过本体建模的方法分别对产品制造工艺、设备次序关系和数据属性特征等内容的描述,提高有序化结构特征和数据属性特征在表达上的灵活性和扩展性,并在此基础上,提出了一种基于工艺特征的自动化生产线本体管理及服务编排系统框架,作为本体模型和相应方法在具体生产制造场景中应用和实现的基础。 然后,本文基于生产过程是按照一定顺序完成的特点,借鉴计算机数据结构中链表的思想和原理,构建了一种具有工艺特征描述的本体模型。除了对常规的物理实体进行描述以外,工艺特征本体中增加了对生产制造过程中有序化结构特征的描述。其中,有序化的结构特征,主要包括设备之间的物理位置次序关系和产品制造工艺的顺序关系两方面内容。这样的表达方式,增强了有序化特征表达的灵活性和扩展性,使得有序化特征的改动能够像链表的操作一样方便快速,有效降低了系统重新配置的代价。 在工艺特征本体模型的基础上,本文提出了一种服务匹配和编排方法。该方法在工艺设计时不再依赖于BPEL流程规范,而是基于有序化结构特征的本体描述,以决策优化求解的过程,对产品的工艺路线和加工步骤进行编排。为了适应多分支生产路径的生产制造情景,该方法将生产线视为计算机数据结构中的树形结构,借鉴递归算法的思想和原理,采用深度优先的搜索方式,实现服务的查找和编排。 为了促进数据的共享和重用,让数据更加智能,能够适用于更广泛的应用范围,本文构建了一种数据属性本体模型,通过对数据属性和特征的语义化描述,使得数据属性独立于具体应用而存在。在此基础上,本文提出了一种感知数据的语义化标注方法。该方法可以在数据访问的过程中,根据具体应用的需求,赋予数据特定的物理意义。这样的方式,不但实现了数据特征的统一化表示,解决了数据异构的问题,提高了数据的可重用性,而且赋予了实时数据更灵活和更丰富的物理意义,提高了数据的智能性,可以适用于更广泛的应用场景,为生产运行过程中应对变化或异常情况时的情景响应和处理提供了更丰富和准确的分析判断依据。 在工艺特征本体和数据属性本体模型的基础上,本文提出了一种面向生产制造过程的情景响应处理方法。该方法能够基于设备服务的状态及变化,对产品的生产过程做出动态的响应和调整,以批量化的生产能力,完成个性化产品的生产。该方法使用了多线程技术和线程同步技术,解决服务调用冲突和服务状态同步更新冲突等问题,以满足多产品并行生产的需要。在生产运行的过程中,产品的生产路径不是预先设定好的,而是基于生产线上设备的状态及变化,由软件系统自主进行选择和调整。 最后,本文以汽车模型装配生产线为实验模拟对象,开发了一个软件实验模拟系统,实现和验证了本文提出本体模型和方法在特定生产场景中的具体应用。

Other Abstract

With the rapid development of the Internet and the maturity of information and communication technology, the modular idea and the characteristics of openness and cross platform provide a feasible technical support for solving the reconfiguration problem of the manufacturing system and improving the flexibility and adaptability of the manufacturing system model. In the field of industrial manufacturing, advances have recently been made toward applying service-oriented approaches or agent technology to consider production equipments as discrete entities similar to web pages. These approaches overcome the limited reconfiguration in the traditional PLC design and facilitate the interconnection and interoperability of heterogeneous production equipments by web service or agent abstraction and encapsulation of the modularization of production equipment with ontology description. They realize resource automatic discovery, invocation and orchestration based on BPEL, with the aim of improving the flexibility and adaptability of the manufacturing system and reducing the reconfiguration cost of the manufacturing system. However, BPEL is a business process execution language, which expresses the manufacturing process of a production object in the way of a whole process. It has strong structure, limited flexibility and expansibility. When the manufacturing process of products changes, the whole BPEL process needs to be redesigned manually. The process of BPEL designing, deploying and compiling is very complicated, which causes a heavy cost to the reconfiguration of the system. In addition, the process of service matching and orchestration based on BPEL is similar to searching and matching web pages on the Internet. It is not only time-consuming, but also involves the problems of recall and precision, which have not been studied in depth in these approaches. Therefore, in order to make the ordered structural features more flexible and scalable in expression, enabling the manufacturing system to orchestrate the manufacturing process based on the product features flexibly for manufacturing the customized production of personalized products by mass production capacity, and to facilitate the understanding and analysis of real-time data in production, enabling the manufacturing system to make responses and adjustments to production processes dynamically based on scenarios, this paper presents an ontology model architecture for production line based on process feature and an ontology management and service orchestration system framework for automatic production line, sets up a process feature ontology and a data attribute ontology, and presents an ontology-based service matching and orchestration method and an ontology-based situational response and processing method on the basis of existing researches. It improves the flexibility and extensibility of the ordered features of production and data attribute features by the description of the manufacturing process knowledge, the relative position of equipments and the characteristics of the data attributes in the ontology model. It also promotes the accuracy of machine's understanding of ordered features and information meanings by the accuracy and non ambiguity expression in the ontology model for providing a technical support for the intelligent response and processing of the manufacturing system based on the situation. On the basis of the transformation of the manufacturing system from the hierarchical structure to the flat structure, this paper realizes continuous transformation to ordered structure. First, this paper presents an ontology model architecture for production line based on process feature, which makes the features of ordered structure and data attributes more flexible and extensible in expression by the description of the manufacturing process knowledge, the relative position of equipments and the characteristics of the data attributes in the ontology model. On the basis of the ontology model architecture, this paper presents an ontology management and service orchestration system framework, which is the basis for the application and implementation of the ontology model and the corresponding methods in the specific manufacturing scene. Then, based on the characteristics of the production process in a certain order, this paper learns from the idea and principle of chain table in computer data structure to set up an ontology model with process feature description, named Process Feature Ontology or PFO for short. The process feature ontology describes not only regular physical entities, but also the features of ordered structure in manufacturing. The features of ordered structure mainly include two parts. One is the order relationship of physical location between equipments, and the other is the sequential relationship of the production process. This description approach enhances the flexibility and extensibility of ordered structure features and makes the modification of ordered structure features as convenient and fast as the operation of the chain table. It effectively reduces the cost of system reconfiguration. On the basis of the process feature ontology, this paper presents a ontology-based service matching and orchestration method. This method makes the process design no longer dependent on the BPEL process specification, but ontology-based description of ordered structural features. It takes the process of optimized solution for decision to orchestrate the process route and processing steps of the product. To adapt to multiple branching paths in the production process, the method regards the production line as the tree structure in computer data structure, learns from the idea and principle of recursive algorithm, and uses the traversal method in depth-first order to implement the search and orchestration of services. In order to promote data sharing and reuse and make the data smarter and useful for a wide range of applications, this paper sets up a data attribute ontology. The data attribute ontology is used to describe data attributes semantically and makes them independent of the specific application. On the basis of the data attribute ontology, this paper presents a semantic annotation method for sensing data. This method can give data specific physical meaning according to the requirements of specific applications in the process of data access. The ontology and the method not only represent data characteristics in a unified way to solve the problems of heterogeneous data and to improve the data reusing, but also give real-time data more flexible and richer meaning, which makes the data smarter and useful for a wide range of applications. They provide more abundant and accurate basis for analysis and decision-making of situational response and processing in the face of changes or abnormal situations in the manufacturing process. On the basis of the process feature and the data attribute ontologies, this paper presents a ontology-based situational response and processing method for manufacturing process. This method makes dynamic response and adjustment to the manufacturing process of the product based on the state and changes of the equipment services, and completes the customized production of personalized products by mass production capacity. In order to satisfy the needs of multiple products peer production, this method uses multi-threading and thread synchronization technologies to solve the problems of conflicts in service invocation and synchronous updates to service status. In the process of production, the production path of the product is not pre-determined, but is chosen and adjusted autonomously by the software system based on the state and state change of devices on the production line. Finally, this paper takes an assembly production line for model cars as an experimental simulation object, and develops a software experimental simulation system to implement and verify the application of these ontologies and methods presented in this paper in a specific production scenario.

Language中文
Contribution Rank1
Document Type学位论文
Identifierhttp://ir.sia.cn/handle/173321/23642
Collection工业控制网络与系统研究室
Recommended Citation
GB/T 7714
施昭. 基于本体的自动化生产线建模方法及其应用[D]. 沈阳. 中国科学院沈阳自动化研究所,2018.
Files in This Item:
File Name/Size DocType Version Access License
基于本体的自动化生产线建模方法及其应用.(2936KB)学位论文 开放获取CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[施昭]'s Articles
Baidu academic
Similar articles in Baidu academic
[施昭]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[施昭]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.