Scheduling of field service resources in cloud manufacturing based on multi-population competitive-cooperative GWO

Abstract

Cloud manufacturing (CMfg) is a new networked manufacturing mode based on the big data and the Internet of Thing technologies, which can dynamically and flexibly allocate the manufacturing resources on demand. All the current researches and applications on CMfg focused on the factory manufacturing schema, the field manufacturing schema has not been concerned. Field manufacturing refers to the geographically dispersed manufacturing resources are arranged to different locations specified by customers to execute their manufacturing tasks, such as the assembly, measurement and maintenance of large equipments, this type of manufacturing is getting more and more applications with the increase of product complexity and the trend of manufacturing industry servicizing. At present, the massive complicated field manufacturing tasks bring urgent needs on the dynamic organization and global planning of the dispersed manufacturing resources, which can be achieved by the network technology. To this end, this paper studies the scheduling of field manufacturing resources in CMfg environment. Firstly, the detailed process of the field service resource scheduling in CMfg environment (FSRS-CMfg) is designed based on analyzing the practical field manufacturing process. After presenting the assumptions and constraints of field manufacturing, the optimization model for the FSRS-CMfg problem is established, in which the overall quality of service (QoS) is used as the optimization objective and its evaluation indicators and aggregation method are elaborately designed. Then, to solve the above optimization model, the encoding and decoding methods are designed and the discrete search operators are developed. For improving the solving speed and accuracy, a multi-population competitive-cooperative grey wolf optimizer (MPCCGWO) is built to avoid trapping in local optimum. Finally, two experiments are carried out to verify the correctness and effectiveness of the proposed model and algorithm. Results show that the overall QoS of the field manufacturing resource scheduling scheme can be markedly improved based on the FSRS-CMfg model, and MPCCGWO possesses higher convergence precision than the commonly-used intelligent optimization algorithms without obvious increase of time consumption, so it is suitable for solving the problems with large solution spaces such as the FSRS-CMfg problem.