**6. Discussion**

In order to improve the intelligence level of safety assessment in the process of structural construction, this study proposed a digital twin modeling method. In the whole research process, the components and environmental information were captured based on the sensing equipment in the physical space. In the virtual space, three-dimensional laser scanning was carried out at each construction step to form the point cloud model of the actual construction process. The geometric model of the structure was corrected by extracting the key node coordinates in the point cloud model. The physical model with high fidelity was finally established to realize the interactive mapping between the virtual space and physical space. The mechanical properties of the structure can be accurately analyzed by the physical model. The conversion probability of the structural state before and after the construction stage of the structure can be obtained. According to the transition probability, the safety state of the structure in the next stage can be predicted by the state of the current structure. In view of the unsafe state, the reasons were traced by the maintenance modeling and the corrective measures were proposed. The feasibility of the measures was analyzed in the physical model. Finally, the construction process of the site was guided to ensure that the structure is in a safe state at all construction stages. Based on the research of the theoretical method, it was applied in the construction process of the wheel–spoke cable truss. The effectiveness of the proposed method was verified by the experimental results, and it improved the intelligence level of construction. At the same time, the safety intelligent assessment methods also provide new ideas and methods for the development of intelligent monitoring of symmetrical structures.
