An application-oriented digital twin framework and the multi-model fusion mechanism

The digital twin (DT) technology facilitates the complete lifecycle management of equipment by integrating physical and virtual spaces through data mapping. Many narrative DT frameworks and modeling methods have been proposed. However, the heterogeneous processes and methods of applying these DT frameworks in different objects and different scenarios of manufacturing restricts the function and promotion of DT. Given that there are some existing discussions on narrative DT frameworks, this paper proposes an application-oriented DT framework that integrates information models, principle models, and field models. Then, the unified DT application process is discussed. The mechanism of how to fuse the multi models for typical applications in evaluation, prediction, and optimization are elaborated in detail respectively. Finally, the proposed framework and application process are validated through two DT models: vehicle wheel polygonal diagnosis digital twin and bogie frame manufacturing optimization digital twin. The correctness and feasibility of the proposed approach is demonstrated through these case studies.