A Production Scheme Design Method for Cloud Manufacturing Service Systems With Resource Uncertainty Using Transition-Rewritable Petri Nets

Cloud manufacturing (CMfg) service systems can provide solutions for large-scale personalized customization. Service composition optimal selection, which is in charge of the selection and allocation of cloud services to multiple tasks, is of great importance for providing significant improvement of CMfg service systems with respect to performance measures such as time and cost. To increase resource utilization and decrease cost consumption, we develop a design method of production schemes for CMfg service systems. Specifically, we define a formalism named transition-rewritable Petri nets (RTPNs) to model and reconfigure the multi-task manufacturing processes in environments allowing resource uncertainty. The reconfiguration process is described by the rewritable rules of RTPNs. On this basis, the performance of the reconfiguration of the manufacturing process is analyzed. In addition, we establish a production strategy considering reconfiguration to obtain the optimized service compositions in terms of time, cost, quality, and reliability. Theoretical comparisons highlight the distinctions between our method and the others in the existing literature. Moreover, the results of simulation experiments and comparative studies show that the method proposed in this paper can obtain better schemes with the lower total completion time and total production cost than a closely-related one presented in a recently published paper. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i>—The uncertainty of resources exists when using cloud manufacturing service system based on distributed resources to complete manufacturing tasks. However, for the large-scale orders, the resource uncertainty is rarely considered by existing work in the literature. This paper proposes a production strategy to ensure that the utilized schemes are always feasible and optimal despite resource uncertainty. A type of Petri nets is defined to model the manufacturing process for the purpose of reconfiguration of the production schemes when the some services change due to resource alterations. Total completion time and total cost are significant considerations. A problem decomposition based genetic algorithm is employed to generate production schemes, which can be easily implemented.