Digital twin-based cyber-physical system for intelligent monitoring and predictive operation of long-span high-speed railway arch bridges

The dynamic interactions between high-speed trains and long-span railway bridges pose significant challenges for the safe and efficient operation and maintenance of these structures. This study presents a digital twin-based cyber-physical system (CPS) designed for real-time monitoring and predictive management of a long-span high-speed railway bridge. The CPS incorporates a train-bridge coupled dynamics model, simulations of train-bridge interactions, a machine learning-powered prediction module, and a fuzzy logic-based decision-making system. Field train load tests were conducted to validate the system’s effectiveness. Results demonstrated that the machine learning model achieved a prediction accuracy of 91 % and reduced computational time by 99.6 % compared to traditional finite element analysis. The fuzzy logic-based decision system provided operational recommendations and accurately identified 92 % scenarios requiring speed restrictions. This research highlights the potential of the CPS to enhance structural health monitoring, improve operational safety, and enable predictive maintenance for complex bridge systems.