Real-time structural health monitoring of bridges using convolutional neural network-based loss factor analysis for enhanced energy dissipation detection

This study introduces a novel method for real-time structural health monitoring (SHM) of bridges using a convolutional neural network (CNN) model that leverages loss factor analysis. As bridge structures deteriorate over time, often accelerated by operational loads, maintaining structural integrity and safety becomes critical. The proposed approach utilizes the concept of a loss factor, which represents the process of energy dissipation across different vibration states, as a key indicator of structural health. This factor is computed from vibration energy spectra, which include amplitude and frequency components, processed through the CNN model for high sensitivity to structural changes. The results demonstrate that the energy dissipation of the bridge during operation can be categorized into signals from three distinct sources: structural responses, defects-related indicators, and noise interference. By monitoring variations in the loss factor over time, the model effectively identifies early signs of structural deterioration, which is critical for timely maintenance interventions. The study also highlights the adaptability to different load conditions and environmental factors, ensuring robust performance in various operational scenarios. The findings underscore the potential of the CNN model to transform SHM practices by enhancing early defect detection, supporting preventive maintenance, and ultimately extending the lifespan of bridge infrastructure.