Computational intelligence model for predicting the compressive strength of FRP-confined concrete column

Fiber reinforced polymer (FRP) wrapping technology is commonly used to enhance the compressive strength (CS) of reinforced concrete (RC) members. Accurate prediction of the compressive strength of FRP-confined concrete columns is crucial for optimizing structural design and helps reduce the time and costs associated with physical testing. Although existing literature and codes have provided corresponding theoretical calculation formulas, the determination of the estimated parameters in these formulas is primarily based on experimental data and engineering experience, resulting in low prediction accuracy. While traditional data-driven models can consider the influence of various factors on the compressive strength of FRP-confined concrete columns, these models often lack clear physical foundations and fail to provide explicit mathematical expressions with engineering significance, making them inadequate for practical engineering needs. This work proposes a hybrid modeling framework that integrates mechanical theory with data-driven methods, aiming to strike a balance between prediction accuracy and physical interpretability. By incorporating additional key influencing factors and a residual learning mechanism, an efficient model is developed for predicting the compressive strength of FRP-confined concrete columns. Ultimately, an expression for the compressive strength of FRP-confined concrete columns, considering multiple factors, is proposed, providing theoretical support for the performance evaluation and design of FRP-strengthened concrete columns.