Integrated Radar and Data-Driven Framework for Modal Analysis and Operational Assessment of High-Speed Railway Bridges

The rapid expansion of high-speed railway networks has increased the demand for reliable and safe bridge infrastructure. Structural health monitoring (SHM) is essential for ensuring long-term performance and safety of these structures. However, conventional contact-based sensing methods, while foundational, face inherent limitations. Particularly in the context of high-speed railway applications. This study introduces a noncontact SHM framework employing interferometric radar for precise and efficient measurement of bridge vibrations. A hybrid time–frequency analysis method, integrating Robust Empirical Mode Decomposition (REMD) with the Multi-Synchro-Squeezing Transform (MSST), is proposed to accurately extract modal parameters and train-induced vibration characteristics from nonstationary signals. Radar-derived data are further used to estimate operational parameters, including train position, carriage number, and velocity, validated through multi-body dynamic simulations. The framework also investigates double-train encounters, revealing distinct time–frequency interaction patterns. To enhance reliability, a Monte Carlo-based approach combining the Natural Excitation Technique (NExT) and Eigensystem Realization Algorithm (ERA) is developed for robust modal parameter estimation. Overall, the proposed radar-based SHM methodology demonstrates high accuracy, scalability, and noninvasiveness for dynamic monitoring of high-speed railway bridges.