Study on Long-Term Temperature Variation Characteristics of Concrete Bridge Tower Cracks Based on Deep Learning.

Monitoring existing cracks is a critical component of structural health monitoring in bridges, as temperature fluctuations significantly influence crack development. The study of the Huai'an Bridge indicated that concrete cracks predominantly occur near the central tower, primarily due to temperature variations between the inner and outer surfaces. This research aims to develop a deep learning model utilizing Long Short-Term Memory (LSTM) neural networks to predict crack depth based on the thermal variations experienced by the main tower. The efficacy of the LSTM network will be rigorously evaluated, employing multiple temperature input datasets to account for spatial dimensional variations in the data. This methodology is anticipated to enhance the model's accuracy in predicting crack widths. By leveraging the deep learning regression model, precise temperature thresholds for crack formation can be established, facilitating early detection of anomalies in the crack widths of the main tower and providing effective technical solutions for monitoring crack status.