Digital twin-driven structure safety monitoring method for revolving floating cranes based on multi-fidelity surrogate models

The Revolving Floating Crane (RFC) plays a crucial role in offshore lifting operations, making its operational safety a top priority. However, large lifting loads and harsh conditions during manual operations can cause unstable postures or structural failures, potentially leading to serious economic losses and personal injuries. To address these safety concerns, comprehensive monitoring of RFC operations is essential. Thus, this study proposes a digital twin (DT)-driven structure safety monitoring method based on multi-fidelity surrogate model (MFSM). In the implementation, the structural and functional characteristics of actual RFCs are referenced to build an RFC physical entity. By identifying key operational safety indicators, a sensor-based data acquisition module for RFC entity is developed. Relying on the entity, finite element simulations and actual experiments are then conducted to generate multi-fidelity training data of RFC operational safety, which are used to establish corresponding MFSMs. By Unity3D engine, a virtual interaction platform for data fusion is created for safety monitoring, enabling real-time synchronization from the RFC physical entity to DT model. Finally, physical safety monitoring experiments are conducted on RFC lifting operations. Experiment results demonstrate that the proposed method delivers high real-time performance and accuracy, effectively meeting the engineering requirements for RFC lifting safety monitoring.