云制造环境下供应商匹配方法研究

发布时间：2018-12-19 11:45

【摘要】：云计算、互联网、物联网等高新技术的发展带动了传统行业的快速转型。云制造是助力“中国制造2025”计划实现制造业发展得重要手段。对于制造企业供应商选择而言,云制造与传统制造相比拥有更大的选择范围,能够主动实时的匹配个性化需求,使分布的制造资源高度共享。同时,云制造庞大的服务资源为供应商选择带来新的难题,其服务匹配是亟待解决的问题。目前传统的匹配方法缺乏针对性,缺少针对供应商领域匹配方法的研究；同时云制造中存在大量功能相似的供应商,需要根据需求的QoS (Quality of Service)信息进行判断,传统匹配机制缺乏对模糊QoS信息的选择和排序。另外,传统匹配方法多推荐唯一的最优服务,难以满足需求方自主选择的需要。针对传统匹配方法的弊端,本文以云制造环境下的供应商为研究对象,结合本体和模糊QoS提出三阶段匹配方法,具体工作如下：首先,分析云制造及其供应链特点、与传统网络制造的区别,以及云平台的服务匹配流程和方法,找到问题的重点和难点,提出本文要解决的问题。然后,定义服务匹配的内容。设计了包含功能性信息和QoS信息的供应商描述模型。在本体建模研究的基础上定义了供应商服务本体的结构,作为匹配算法的语义支持。接着,设计供应商服务匹配算法,算法分为三个阶段,每个阶段匹配的内容分别是服务的功能性信息、QoS信息和前两个阶段的综合信息。针对功能性信息的资源概念,采用了基于语义相似度的匹配算法,同时为避免了由于表达方式不同造成概念不能匹配到本体模型,从而造成错误或片面匹配结果的问题,引入了属性影响因素。针对QoS信息匹配,考虑了QoS的模糊性和用户偏好,采用优化的FCM (Fuzzy C Means Clustering)算法。匹配方法结果为一定数量的基于综合匹配度的排序结果集合,满足需求方自主挑选供应商的需求。最后,实现了服务匹配系统并对匹配算法进行实验分析。通过实例验证了匹配方法的有效性,能较好的满足现实需求。并对比分析了查全率和查准率,同时验证了优化的算法提高了服务匹配的效率。本研究为探索如何解决云制造环境下的供应商匹配问题提出了一种新思路。
[Abstract]:Cloud computing, Internet of things, Internet of things and other high-tech development has driven the rapid transformation of traditional industries. Cloud manufacturing is an important means to help the "made in China 2025" plan to realize the development of manufacturing industry. For supplier selection of manufacturing enterprises, cloud manufacturing has a larger selection range than traditional manufacturing, and can actively and real-time match personalized requirements, so that distributed manufacturing resources are highly shared. At the same time, the huge service resources of cloud manufacturing bring new problems for supplier selection, and service matching is an urgent problem to be solved. At present, the traditional matching methods are lack of pertinence and lack of research on supplier domain matching methods. At the same time, there are a large number of similar suppliers in cloud manufacturing, which need to be judged according to the required QoS (Quality of Service) information. The traditional matching mechanism lacks the selection and ranking of fuzzy QoS information. In addition, the traditional matching methods often recommend the only optimal service, which is difficult to meet the needs of the demand side. Aiming at the disadvantages of traditional matching methods, this paper takes suppliers in cloud manufacturing environment as the research object, combines ontology with fuzzy QoS, and proposes a three-stage matching method. The specific work is as follows: firstly, the characteristics of cloud manufacturing and its supply chain are analyzed. The difference from the traditional network manufacturing, as well as the service matching flow and method of the cloud platform, find out the key and difficult points of the problem, and put forward the problems to be solved in this paper. Then, define the content of the service match. A supplier description model including functional information and QoS information is designed. On the basis of ontology modeling, the structure of supplier service ontology is defined as the semantic support of matching algorithm. Then, the supplier service matching algorithm is designed. The algorithm is divided into three stages. The content of each phase matching is the functional information of the service, the QoS information and the synthesis information of the first two stages. For the resource concept of functional information, a matching algorithm based on semantic similarity is adopted, and in order to avoid the problem that concepts can not match to ontology model because of different expression methods, thus causing errors or one-sided matching results. The influence factor of attribute is introduced. Considering the fuzziness and user preference of QoS, the optimized FCM (Fuzzy C Means Clustering) algorithm is adopted. The result of the matching method is a set of ranking results based on the comprehensive matching degree to meet the demand of the demand side. Finally, the service matching system is implemented and the matching algorithm is analyzed experimentally. The effectiveness of the matching method is verified by an example, and it can meet the practical requirements. The recall rate and precision ratio are compared and analyzed, and the optimized algorithm is verified to improve the efficiency of service matching. This study provides a new way to solve the problem of supplier matching in cloud manufacturing environment.
【学位授予单位】：大连理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：F274

【参考文献】