Integrated Service Matching and Composition for Cloud Manufacturing Platforms

Cloud manufacturing has been recognized as a transformative manufacturing paradigm aiming towards producing highly customized products via sharing distributed manufacturing resources and capabilities. One of the pivotal challenges regarding the practical realization of this idea is the process of matching manufacturing resources with personalized service demands. This problem contains two main aspects: (1) retrieval of functionally similar services to form corresponding service candidate sets, and (2) optimal composition of subtasks according to non-functional quality of service (QoS) indexes. However, almost all the research in the field thus far has focused on tackling each of these dimensions individually which hardly corresponds to actual conditions of cloud manufacturing paradigm. To this end, this dissertation proposes a novel integrated approach that addresses these two problems simultaneously. First, TF-IDF (Term Frequency-Inverse Document Frequency) method coupled with classification algorithms is deployed to retrieve and form the candidate sets. This addresses the oversimplification existing in the literature regarding the predefined service candidate set. Besides, using real-world manufacturing capability data instead of random candidates adds another layer to achieve an even more comprehensive model of the problem. Two novel metaheuristic algorithms were proposed and implemented to effectively solve the service composition and optimal selection problem in the context of cloud manufacturing, especially for large-scale scenarios. The results substantiate the effectiveness of the proposed approaches.