Abstract: Under mass personalization production environments, manufacturing resources within the workshop needed to be reorganized frequently to build new service collaboration chains to complete different manufacturing tasks. In order to improve efficiency and stability of the service collaborative chains, a manufacturing resource aggregation model was established to describe the aggregation relationships of workshop manufacturing resources. Manufacturing resources were encapsulated into workshop manufacturing services based on improved K-means ++ algorithm. Frequent sub-network mining algorithm was used to mine frequently occurring service collaboration chains, and service collaboration chain evaluation was proposed to evaluate and filter the obtained service collaboration chains. Finally, feasibility and effectiveness of the proposed method were verified by case analysis.

Key words: mass personalization, manufacturing service, encapsulation, service collaboration, data mining

摘要： 大规模个性化制造模式下，车间内部的制造资源需频繁进行重组以构建新的服务协作链来完成不同的制造任务。为提高服务协作链的高效性和稳定性，建立了制造资源聚合模型来描述车间制造资源的聚集关系，并基于改进的K-means++算法将制造资源封装成车间制造服务。使用频繁子网络挖掘算法挖掘了频繁出现的服务协作链，并提出服务协作链评价方法来对获取的服务协作链进行评价和筛选。最后通过案例分析对所提方法的可行性和有效性进行了验证。

关键词: 大规模个性化, 制造服务, 封装, 服务协作, 数据挖掘