A multi-process parallel clustering algorithm for resource reconfiguration in cloud manufacturing

Resource reconfiguration can integrate and optimize manufacturing resources (MRs) into various resource sets, enhancing overall resource management in cloud manufacturing (CM). However, achieving efficient resource reconfiguration remains difficult due to the large data volumes, heterogeneity, and implicit relationships among MRs in CM. To improve the practicality of MR reconfiguration in cloud manufacturing (CM), a new multi-process parallel clustering algorithm (MPPCA) is proposed to categorize the multi-source heterogeneous MRs into distinct resource sets based on their functional characteristics. MPPCA can divide MRs into multiple processes and cluster them parallel to form several subclasses. Finally, similar subclasses are merged, and MRs with implicit relationships are grouped into the same resource set. In order to ensure the accurate calculation of the subclass radius during the merging process, a skewness indicator is introduced. The indicator is used to evaluate the irregular shape of the data set during the clustering process to ensure that the subclasses after division are quasi-circular. Additionally, merge criteria are proposed to merge similar resource subclasses effectively, exploring implicit relationships between MRs. The feasibility of the proposed method is validated through experiments using MR data from laboratories and artificial datasets, demonstrating that MPPCA successfully achieves resource reconfiguration.