Under the influence of Industry 4.0, manufacturing resources (MRs) across regions and enterprises realize ubiquitous interconnection and real-time data acquisition via communication network and internet of things, which makes it possible for distributed MRs to provide appropriate production services according to changeable market demand. Therefore, social manufacturing (Social Mfg) system came into being. Under the Social Mfg environment, a flexible configuration method of distributed MRs is put forward in this paper, which is qualified to quickly respond to order demand and production disturbance. Firstly, superior manufacturing services are selected from the resource pool of the platform according to the types of subtasks. Secondly, high-quality service compositions capable of completing a complicated manufacturing task are obtained by NSGA-III algorithm, which meets the requirements of multi-objective optimization in terms of time, cost and quality. In addition, a MRs collaborative network model is constructed based on the historical cooperation data. The Louvain algorithm is employed to detect the manufacturing communities (MCs) from the collaborative network. Finally, the MCs are applied to provide the optimal service schemes for personalized demand at a fast speed, as well as realizing dynamic reconfiguration of MRs under abnormal disturbance. Experimental results demonstrate that the proposed method possesses significant performance in both efficiency and robustness, and makes great contributions to the development of Social Mfg in terms of resource configuration.

在工业4.0的影响下，跨地域、跨企业的制造资源（MR）通过通信网络和物联网实现泛在互联和实时数据采集，使得分布式MR可以根据多变的市场需求提供合适的生产服务. 因此，社会制造（Social Mfg）体系应运而生。在Social Mfg环境下，本文提出了一种分布式MR的灵活配置方法，能够快速响应订单需求和生产扰动。首先，根据子任务的类型，从平台的资源池中选择优质的制造服务。第二，NSGA-III算法得到能够完成复杂制造任务的高质量服务组合，满足多目标优化的时间、成本和质量要求。此外，基于历史合作数据构建了MRs合作网络模型。Louvain 算法用于从协作网络中检测制造社区 (MC)。最后，应用MCs为个性化需求快速提供最优服务方案，实现异常扰动下MRs的动态重构。实验结果表明，所提出的方法在效率和鲁棒性方面都具有显着的性能，并且在资源配置方面为社会制造的发展做出了巨大贡献。