基于时间序列的制造云服务选择研究

发布时间：2018-09-09 12:49

【摘要】：随着“工业4.0”时代的到来,信息技术与工业制造技术高度融合、深度交织产生了新的制造模式——云制造。该模式利用云制造服务平台,按用户的实际需求组织平台中的制造资源,给用户提供各类制造服务,为解决当前我国制造领域存在的不合理利用制造资源而引起资源浪费的问题提供了新的解决思路。因此,如何从云制造平台大量的共享制造资源中选择合适的制造服务成了研究热点。本文以汽车制造厂零部件生产为例,研究了相应的制造商选择问题,通过制造云服务选择方法为汽车制造厂选择最合适的零部件生产制造云服务商,从而使得汽车制造厂最大限度满足自身的制造需求。本文的研究亦可为其他制造企业在制造云服务选择问题上提供参考。首先,本文通过查阅大量相关文献资料,结合制造资源和网络性能两方面,构建了一套具有制造云服务特色的QoS(Quality of Service)评价指标体系。然后,针对制造云服务随时间变化的特性,引入“时间序列”的概念,将制造云服务QoS动态变化情况用时间序列形式表达。考虑到实际应用中由于设备故障等原因情况,会出现QoS观测数据在某些时间序列节点缺失的问题。本文利用基于云服务QoS序列特性的缺失值估计方法预测和填补相应的缺失值,较好地解决了这一问题。其次,本文提出了一种基于主客观综合权重的云服务时间序列选择方法。从基于用户QoS偏好层次的主观权重和基于QoS指标相关性的客观权重两个角度进行考虑,通过结合时间序列QoS模型进行云服务选择,为制造企业提供帮助。最后,通过仿真实验分析表明,该方法在有效解决用户QoS偏好的同时又充分考虑云服务集的QoS指标的数据分布特性,选择结果具有较高的准确性与科学性。
[Abstract]:With the arrival of the "industry 4.0" era, information technology and industrial manufacturing technology are highly integrated, and a new manufacturing model-cloud manufacturing is produced by the deep interweaving of information technology and industrial manufacturing technology. This model uses the cloud manufacturing service platform to organize the manufacturing resources in the platform according to the actual needs of the users, and provides all kinds of manufacturing services to the users. It provides a new way to solve the problem of resource waste caused by irrational use of manufacturing resources in the field of manufacturing in China. Therefore, how to select suitable manufacturing services from a large number of shared manufacturing resources on cloud manufacturing platform has become a hot research topic. In this paper, taking the manufacture of parts and components in automobile factory as an example, the problem of manufacturer selection is studied. Through the method of manufacturing cloud service selection, the most suitable component manufacturing cloud service provider is selected for automobile manufacturing plant. In order to make the automobile factory to meet its own manufacturing needs to the maximum extent. The research in this paper can also provide reference for other manufacturing enterprises in the selection of manufacturing cloud services. Firstly, this paper constructs a set of QoS (Quality of Service) evaluation index system with the characteristics of manufacturing cloud service by consulting a lot of relevant literature and combining manufacturing resources and network performance. Then, according to the characteristics of manufacturing cloud service with time, the concept of "time series" is introduced, and the dynamic change of manufacturing cloud service QoS is expressed in the form of time series. Considering the situation of equipment failure and other reasons in practical application, the problem of missing QoS observation data in some time series nodes will occur. In this paper, the method of estimating the missing values based on the characteristics of cloud service QoS sequences is used to predict and fill the corresponding missing values, which solves this problem well. Secondly, this paper proposes a cloud service time series selection method based on subjective and objective weight synthesis. Considering the subjective weight based on the user QoS preference level and the objective weight based on the correlation of QoS index, the cloud service selection is carried out by combining the time series QoS model to provide help for manufacturing enterprises. Finally, the simulation results show that the method not only solves the user's QoS preference effectively, but also fully considers the data distribution characteristics of the QoS index of the cloud service set, and the selection result is accurate and scientific.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F424;F49

【参考文献】