Journal of Engineering Research

Smart manufacturing systems based on digital twin have become a crucial path to drive the upgrading of the manufacturing industry towards intelligence and autonomy. This paper focuses on the research of real-time optimization and autonomous decision-making, with emphasis on exploring how digital twin achieves dynamic perception, prediction, and feedback adjustment of the production process through virtual-real mapping. The study proposes a system framework centered on multi-source data fusion, virtual model-driven, and adaptive algorithm optimization, which enables real-time adjustment of resource allocation and intelligent optimization of process flows in production. By introducing an autonomous decision-making mechanism, the system can respond quickly in complex environments and improve the robustness and flexibility of the manufacturing system. The research results of this paper show that the smart manufacturing system based on digital twin not only has significant advantages in improving production efficiency and product quality but also provides a feasible path option for the construction of future smart factories.

Digital twin; Smart manufacturing; Real-time optimization; Autonomous decision-making