Fiber reinforced polymer (FRP) has the strengths of lightweight, high strength, and easy construction, and is extensively used in the reinforcement and strengthening of civil engineering structures. To address the issues of inconsistent prediction models, cumbersome calculation processes, and limited precision in predicting the flexural bearing capacity of FRP-concrete materials, this paper designs a prediction model based on Convolutional Neural Network (CNN), Attention Mechanism, and Long Short Term Memory Network (LSTM) (CNN-Attention-LSTM). CNN extracts features from collected data, while LSTM learns better from time series data. Based on the CNN-LSTM model, attention mechanism is coupled to highlight the impact of features on input performance, thereby enhancing computational speed. Taking a certain FRP-concrete face rockfill dam as an example, the predicted bending bearing capacity of the model was compared with the prototype monitoring data through CNN-Attention-LSTM model calculation, and then compared with the prediction results of the traditional model. The results indicate that CNN-Attention-LSTM model has better predictive performance.
Xin Zheng (Sun,) studied this question.