Abstract: To improve the service composition flexibility and reduce the service energy consumption in the operation of cloud manufacturing services,a seven-dimensional cloud manufacturing service evaluation index system was established,including service execution time,service delivery quality,service cost,availability,reliability,service energy consumption and service composition flexibility.Then,a cloud service composition model with the optimal service quality,minimum energy consumption and maximum service composition flexibility was constructed.Combining the advantages of particle swarm algorithm and clone selection algorithm,and the strategies of Lévy flight and chaotic mapping,a multi-objective particle swarm and clone selection fusion algorithm based on adaptive mesh technology was proposed to solve the constructed model.Finally,the feasibility and effectiveness of the proposed method were verified by comparing with NSGA-II,MOPSO and MOGWO algorithms in benchmarking problems and a mould manufacturing case.

Key words: cloud manufacturing, service composition, service flexibility, service energy consumption

摘要： 为提高云制造服务组合的柔性并降低制造任务执行过程中的运行能耗,建立包括服务执行时间、服务交付质量、服务成本、可用性、可靠性、服务能耗和服务组合柔性七维目标的云制造服务评价指标体系,构建以服务质量最优、能耗最低、服务组合柔性最大为优化目标的云制造服务组合优选模型。在此基础上,基于粒子群算法和克隆选择算法各自的优势,融合莱维飞行、混沌映射等策略,提出基于自适应网格技术的多目标粒子群和克隆选择融合算法对所构模型进行求解。最后,在基准测试问题和模具制造案例中,通过与NSGA-Ⅱ、MOPSO和MOGWO算法进行对比,验证所提方法的可行性和有效性。

关键词: 云制造, 服务组合, 服务柔性, 服务能耗