基于案例推理的虚拟供应链风险评估研究

发布时间：2019-01-07 14:08

【摘要】：随着经济全球化的不断发展,企业面临的竞争范围不断扩大,企业仅凭自己有限的资源很难在市场竞争中取胜,目前,全球商业竞争已发展为供应链的竞争。而传统的供应链模式,因其自身结构的局限性,价值散落在整条供应链的不同环节,使供应链的创新发展受到了限制。因此,很多企业开始跳出这种典型的供应链模式,站在全局的角度对整条供应链进行管理,利用外部资源优势,形成一条虚拟的供应链,而将散落在各个环节的价值归于一身,形成自身的竞争优势。虚拟供应链作为一种新的管理机制,也存在其特定的风险,任何一个环节出现问题都可能影响整条供应链的正常运作,因此,研究虚拟供应链的风险评估具有很重要的现实意义。本文首先对案例推理和虚拟供应链风险的研究现状和理论进行综述,确定案例推理方法适合于进行虚拟供应链风险评估,并提出把案例推理技术引入到虚拟供应链风险评估的研究中。在对虚拟供应链风险类型及影响因素研究的基础上本文构建了虚拟供应链风险评估指标体系,并定义了各类风险指标的含义。采用框架表示法对虚拟供应链风险案例进行表示,结合虚拟供应链风险领域知识的特点,建立了虚拟供应链风险案例的顶层框架和子框架。另外,实现案例库的构建,详细规划和设计了案例库结构图和表结构,并采用知识引导法建立案例索引,以风险类别为主索引,并设置副索引和备选索引,随着案例库的不断更新和学习,逐步完善案例库的索引机制,以加强对案例库的管理,提高案例检索的效率。在案例检索方面,基于广义梯形模糊数的概念,本文改进并提出一种新的混合相似度度量方法,克服了大多数方法不能有效处理模糊数的缺陷,同时,设计了案例相似度检索机制,为案例推理模型的实现打下基础。在案例修正方面,本文根据案例检索的三种可能结果,给出了不同的案例修正措施,尤其是检索到K个近似案例的情况下,采用计算匹配率和准确度的方法进行修正,以提高系统决策准确率。关于案例学习,本文在传统的案例学习的基础上,提出建立积淀案例库的概念,通过属性约简的方法形成积淀案例,简化案例检索的特征指标,提高案例库的学习能力和检索效率。最后,通过实例研究了系统进行风险评估的过程,证明了系统的可行性及有效性。本文基于案例推理的应用和虚拟供应链风险评估作了新的尝试,提高了问题求解的实时性和智能化水平,为虚拟供应链风险的评估和防范,提供了一种新的研究思路和解决方法。
[Abstract]:With the development of economic globalization, the scope of competition faced by enterprises is constantly expanding. It is very difficult for enterprises to win in the market competition only with their limited resources. At present, the global commercial competition has developed into the competition of supply chain. Because of the limitation of its own structure, the value of the traditional supply chain is scattered in different links of the whole supply chain, which limits the innovation and development of the supply chain. Therefore, many enterprises begin to jump out of this typical supply chain mode, manage the whole supply chain from the perspective of the overall situation, take advantage of external resource advantages, form a virtual supply chain, and put the value scattered in each link into one. Form its own competitive advantage. As a new management mechanism, virtual supply chain also has its specific risks. Any problems in any link may affect the normal operation of the whole supply chain. It is of great practical significance to study the risk assessment of virtual supply chain. In this paper, the current situation and theory of Case-Based reasoning (CBR) and Virtual supply chain (VSC) risk are reviewed firstly, and the Case-Based reasoning (CBR) method is determined to be suitable for risk assessment of Virtual supply chain A case-based reasoning (CBR) technique is proposed to study the risk assessment of virtual supply chain. Based on the study of the risk types and influencing factors of virtual supply chain, this paper constructs the risk assessment index system of virtual supply chain, and defines the meaning of various risk indicators. The framework representation method is used to represent the risk case of virtual supply chain. Combining the characteristics of risk domain knowledge of virtual supply chain, the top-level framework and sub-framework of the risk case of virtual supply chain are established. In addition, the case base structure diagram and table structure are planned and designed in detail, and the case index is established by knowledge leading method. The risk category is the main index, and the secondary index and alternative index are set up. With the continuous updating and learning of the case base, the index mechanism of the case base is improved step by step in order to strengthen the management of the case base and improve the efficiency of case retrieval. In case retrieval, based on the concept of generalized trapezoid fuzzy number, this paper improves and proposes a new mixed similarity measure method, which overcomes the defect that most methods can not deal with fuzzy number effectively. A case similarity retrieval mechanism is designed to lay the foundation for the realization of case-based reasoning model. In the aspect of case correction, according to the three possible results of case retrieval, this paper gives different case correction measures, especially when K approximate cases are retrieved, the method of calculating matching rate and accuracy is used to correct them. In order to improve the system decision accuracy. As for case learning, based on the traditional case study, this paper puts forward the concept of establishing the accumulated case base, forms the accumulated case by attribute reduction, and simplifies the characteristic index of case retrieval. Improve the case base learning ability and retrieval efficiency. Finally, the process of risk assessment is studied through an example, and the feasibility and effectiveness of the system are proved. In this paper, the application of case-based reasoning and the risk assessment of virtual supply chain are made a new attempt to improve the real-time and intelligent level of problem solving, which is the risk assessment and prevention of virtual supply chain. A new research idea and solution are provided.
【学位授予单位】：东华大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F274

【参考文献】