Artificial Intelligence–Based Structural Health Monitoring of Aging Reinforced Concrete Bridges Using Sap2000 Simulation and Vibration Data Analysis

The study presents a hybrid Artificial Intelligence–based Structural Health Monitoring (SHM) framework for aging reinforced concrete highway bridges, integrating SAP2000 finite element simulations with vibration data analysis. Two multi-span bridges—an older span and a newly strengthened span—were instrumented and tested under controlled vehicular loading using 35-ton and 45-ton trucks at varying speeds between 40–80 km/h. Field measurements of acceleration and deflection were obtained through high-frequency sensors and Fast Fourier Transform (FFT) analysis and compared with numerical simulations. Results indicated that bridge response is highly sensitive to vehicle speed and load, with Dynamic Amplification Factors (DAF) ranging from approximately 1.00–2.35 for the old bridge and 1.00–1.89 for the new bridge. The integration of experimental data with a Convolutional Neural Network (CNN) model significantly improved predictive accuracy, achieving R² values above 0.80 when benchmarked against FEM outputs and approximately 0.99 when compared with field data. The hybrid CNN–DAF framework effectively enhances the reliability of structural response prediction, maintaining modal consistency while addressing the limitations of traditional FEM analysis. This approach provides a robust, data-driven solution for continuous bridge health monitoring, enabling proactive maintenance, extended service life, and improved safety of critical transportation infrastructure.