Composite bridges are typically exposed to temperature variations due to heat radiation, conduction, and convection. Temperature affects the modal parameters of bridges, hindering the application of damage detection methods based on the dynamic properties of bridges. In this study, the effects of temperature on the natural frequencies of composite bridges were investigated experimentally and numerically to derive a basis for separating temperature effects from the natural frequencies. A temperature-controllable girder specimen was developed for modal testing. Additionally, finite element (FE) analysis was conducted to analyze the effects of temperature. The FE analysis results were validated by comparing them to the static response results of the test specimen. The results of the experiments and FE simulations verified that temperature variation can affect the material properties, particularly the modulus of elasticity, of a composite girder, consequently influencing its natural frequency. Based on the tests and simulations, a linear relationship between the temperature and the natural frequency was proposed to remove the temperature effects from the natural frequency.

The dynamic characteristics of bridges are not only influenced by structural damage but also by ambient environmental variations. If environmental factors are not incorporated into the detection algorithm, they may lead to false positives or false negatives. In recent years, vibration-based damage detection methods have gained significant attention in structural health monitoring (SHM), particularly for assessing structural integrity under varying temperature conditions. This study introduces a deep-learning framework for identifying damage in composite bridges by utilizing both time-domain and frequency-domain vibration signals while explicitly accounting for temperature effects. Two deep learning models—Convolutional Neural Network (CNN) and Artificial Neural Network (ANN)—were implemented and compared. The effectiveness of the proposed damage identification approach was evaluated using an experimental dataset obtained from a composite bridge structure. Furthermore, statistical evaluation metrics—including accuracy, precision, recall, F1 score, and the ROC curve—were used to compare the damage detection performance of the two deep learning models. The results reveal that the CNN model consistently outperforms the ANN in terms of classification accuracy. Moreover, frequency-domain analysis was shown to be more effective than time-domain analysis for damage classification, and integrating temperature data with vibration signals improved the performance of all model architectures.