›› 2021, Vol. 27 ›› Issue (6): 1780-1798.DOI: 10.13196/j.cims.2021.06.023

Previous Articles     Next Articles

Cloud manufacturing service composition optimization based on reliability and credibility analysis

  

  • Online:2021-06-30 Published:2021-06-30
  • Supported by:
    Project supported by the National Natural Science Foundation of China and the Royal Society of Edinburgh,China (No.51911530245),the China Scholarship Council,China(No.[2020]1509),the GUES Scientific Research Foundation for Advanced Talents,China(No.G2018009),and the Natural Science Research Foundation of Guizhou Provincial Education Department,China(No.[2019]158).

基于可靠性可信性分析的云制造服务组合优化

李永湘1,2,姚锡凡2+,刘敏2   

  1. 1.贵州工程应用技术学院机械工程学院
    2.华南理工大学机械与汽车工程学院
  • 基金资助:
    国家自然科学基金委员会与英国爱丁堡皇家学会合作交流项目(51911530245);国家留学基金管理委员会资助项目(留金美[2020]1509号);贵州工程应用技术学院高层次人才科研启动资助项目(院科合字G2018009);贵州省教育厅自然科学研究项目(黔教合KY字[2019]158)。

Abstract: In view of the impact of unstable manufacturing entity reliability and service reputation on the new-era manufacturing industry,the reliability and credibility of cloud manufacturing service were analyzed.Combining service reliability and credibility,composition complexity and synergy with execution time and cost,a new service quality evaluation model was constructed.The performance of service composition scheme was evaluated by calculating weighted relative deviation,and an Entropy Enhanced Particle Swarm Optimization (EEPSO) algorithm was proposed.The normal cloud model was introduced to improve the algorithm,which improved its global search ability in the early stage and local search accuracy in the later stage.Taking lifting assembly robot manufacturing task as an example,the validity of the proposed multi-objective optimization model for cloud manufacturing service composition and the feasibility of EEPSO algorithm were verified.Case studies showed that EEPSO had faster convergence speed and better comprehensive performance compared with SGA,CSBHC and so on.

Key words: cloud manufacturing, entropy enhanced particle swarm optimization, service composition, service reliability, service credibility

摘要: 针对不稳定的制造实体可靠性和服务信誉给新时代制造带来的影响,分析了云制造服务可靠性和可信性,将服务可靠度和可信度、组合复杂度和协同度与执行时间和费用相结合,构建了一种新的服务质量(QoS)评价模型;并通过加权相对偏差评价服务组合性能,提出一种熵增强粒子群优化算法(EEPSO),再引入正态云以提高算法前期全局搜索能力和后期局部寻优精度。以举升装配机器人制造任务为例,验证了优化模型有效性和EEPSO算法可行性。结果表明,与标准遗传算法(SGA)、混合布谷鸟算法(CSBHC)、粒子群优化算法(PSO)、云熵遗传算法(CEGA)这4种算法相比,EEPSO具有更快收敛速度和更好综合性能。

关键词: 云制造, 熵增强粒子群优化算法, 服务组合, 服务可靠度, 服务可信度

CLC Number: