Abstract: To realize full sharing, exploration and matching of manufacturing resource and to provide optimal cloud service under cloud manufacturing, technologies of resource modeling and clustering were introduced into the stage of virtual resource pool construction and a framework concerning virtual resource modeling and optimization method was proposed. A bilayer model based on XML data and OWL semantic was employed to describe virtual resource modeling and digitalization of cloud pool. Based on cost, efficiency, quality and credit of service, a mathematical multi-objective optimization model was constructed. The particle swarm optimization algorithm was used to solve the optimization problem, and a Pareto optimal solution was obtained to form instantiated virtual manufacturing cell. An instance from manufacturing practice was used to demonstrate that the proposed framework could improve the sharing, exploration and matching performance of cloud pool's virtual resource efficiently. Its validity in engineering application was also verified.

Key words: cloud manufacturing, virtual resource pool, virtual cell, particle swarm optimization algorithm, modeling

摘要： 为实现云制造环境下制造资源的全面共享、发现和匹配，提供给用户理想优质的云制造服务，在虚拟资源池构建阶段引入资源建模与聚类技术，提出一种适宜云制造模式的虚拟资源的建模与优化组合方法。采用基于可扩展标记语言数据与Web本体描述语言语义的双层模型，对云池的虚拟资源协同进行建模；建立基于服务成本、服务效率、服务质量和信誉度的多目标优化数学模型，采用协同多目标粒子群算法求解该优化问题，获得最优的虚拟制造单元。以生产实践中开发的原型系统为例，验证了所提方法能够有效提高云池虚拟资源的共享、发现和匹配性能，也验证了其在工程应用中的正确性。

关键词: 云制造, 虚拟资源池, 虚拟单元, 粒子群优化算法, 建模