摘要
随着云制造服务模式的兴起,云池中存在大量云服务并掺杂有虚假的云服务,恶意竞争的虚假云服务伤害了云服务提供商的权益,云用户也很难判别真实且优良的云服务。针对该问题,提出基于蒙特卡洛模拟估计的超体积算法来识别和剔除虚假云服务。在剔除虚假云服务的基础上,提出面向云服务可信度评估的基于遗传算法的模糊c均值聚类算法对云服务聚类并进行可信度评估,利用由层次分析法和熵值法构成的制造云服务组合赋权综合评价法,根据可信度等级结合综合评价值对云服务进行优选。以某工程机械交易云平台作为研究对象,验证了所提算法的有效性,算法能有效识别虚假云服务,确定云服务的优先顺序,并选出最佳的云服务。
With the development of cloud manufacturing service model,a lot of cloud services mixed with fake cloud services exist in the cloud pool,thus the rights of cloud service providers are damaged because of the fake cloud services that aim to the vicious competition,and it is difficult for the cloud service users to choose real and optimal cloud service.To solve this problem,a hypervolume algorithm based on Monte Carlo simulation for eliminating fake cloud service was proposed.After eliminating the fake cloud service,a fuzzy C-means clustering algorithm based on genetic algorithm for the credibility assessment of cloud service was proposed and used to cluster the cloud service and evaluate the credibility of the clustering center.The comprehensive evaluation method with combination weighting that was composed of Analytic Hierarchy Process(AHP)and entropy value method was introduced to evaluate the cloud service for obtaining the comprehensive values,and the cloud services were optimally chosen based on the credibility rating and the comprehensive values.Through a machinery trading cloud platform,the effectiveness of the proposed algorithm was verified.The results showed that algorithm could effectively identify the fake cloud services,determine the priority of the cloud services and chose the best cloud service.
作者
江泽豪
伊德景
朱光宇
JIANG Zehao;YI Dejing;ZHU Guangyu(School of Mechanical Engineering&Automation,Fuzhou University,Fuzhou 350116,China)
出处
《计算机集成制造系统》
EI
CSCD
北大核心
2020年第8期2020-2029,共10页
Computer Integrated Manufacturing Systems
基金
工信部2016智能制造应用资助项目(工信部联装(2016)213)
CAD/CAM福建省高校工程研究中心开放基金资助项目(K201704)。
关键词
虚假云服务
蒙特卡洛超体积算法
云服务聚类
可信度评估
综合评价
fake cloud service
Monte Carlo hypervolume algorithm
cloud service clustering
credibility assessment
comprehensive evaluation
