基于双前沿面数据包络分析的云制造服务功能匹配  

Cloud manufacturing service functional matching based on double frontiers DEA

作　　者：张志颖[1] 张红艳 刘茹梦 魏旭光[1] ZHANG Zhiying;ZHANG Hongyan;LIU Rumeng;WEI Xuguang(School of Economics and Management,Hebei University of Technology,Tianjin 300401,China)

机构地区：[1]河北工业大学经济管理学院,天津300401

出　　处：《计算机集成制造系统》2024年第4期1374-1388,共15页Computer Integrated Manufacturing Systems

基　　金：国家社会科学基金资助项目(17BGL195)。

摘　　要：针对云制造环境下功能匹配过程中属性权重需提前确定,进而影响匹配精度的问题,提出了一种基于本体相似度的双前沿面数据包络分析(DEA)匹配方法,该方法根据总体几何平均相似度进行服务排序,并自动生成属性权重。首先,为了提高匹配效率,从基本信息、类型、状态及功能四方面对云制造服务与需求进行描述;然后,引入信息量并改进相似度计算方法,提出基于概念语义距离、信息量和概念层次顺序的本体概念综合语义相似度计算方法;在此基础上,运用双前沿面DEA模型衡量总体效率,以此进行服务排序;最后通过算例证明了该方法的可行性。所提方法为改善云制造服务功能匹配效果提供了参考。In the cloud manufacturing environment,the functional matching process of services and requirements often has the problem that the functional attribute weights need to be determined in advance,which affects the matching accuracy.To solve the problem,a double frontiers Data Envelopment Analysis(DEA)matching method based on ontology similarity was proposed.This method ranked services according to the overall geometric average similarity,and the attribute weights were automatically generated.The cloud manufacturing services and demands were described from the four aspects of basic information,categories,status and functions.Then,information content was introduced and the similarity calculation method was improved,and a new method was proposed to calculate the comprehensive semantic similarity of ontology concepts based on the semantic distance of concepts,the information content and the hierarchical order of concepts.The services were ranked on the basis of the overall efficiency by the double frontier DEA model.Finally,the method feasibility was proved through an example.This paper provided reference for improving the functional matching effect of cloud manufacturing services.

关 键 词：云制造服务 功能匹配 双前沿面数据包络分析 综合相似度 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]