Abstract: The bond performance between Fiber Reinforced Plastics (FRP) bars and concrete is a key factor affecting the mechanical properties of FRP reinforced concrete structures. Currently, there is no efficient and accurate prediction model for the bond strength between FRP bars and multiple types of concrete. Based on this, this study proposes a prediction model for the bond strength between FRP bars and multiple types of concrete that combines Sparrow Search Algorithm (SSA) and Extreme Gradient Boosting (XGBoost) algorithm. A database of bond strength between 348 sets of FRP bars and various types of concrete is established using experimental data from existing literatures. The database is randomly divided into training and testing sets in a 4∶1 ratio, and the SSA-XGBoost algorithm is used to predict the bond strength between FRP reinforcement and various types of concrete. Horizontal comparisons are made with commonly used prediction algorithms such as Random Forest (RF), Support Vector Regression (SVR), and Extreme Gradient Boosting (XGBoost) to verify the reliability and accuracy of prediction results of the algorithm model. SHAP was used to conduct interpretability analysis on the SSA-XGBoost model, revealing the specific contribution of each input parameter to prediction results of the model. The results showed that compared to the three machine learning prediction models SVR, RF, and XGBoost, the SSA XGBoost model has a 1.82%, 2.53%, and 1.31% increase in coefficient of determination (R²) on the training set, respectively. The root mean square error (RMSE) has decreased by 58.09%, 44.73%, and 31.02%, respectively. The mean absolute error (MAE) has decreased by 48.02%, 52.07%, and 38.75%, respectively. The symmetric mean has absolute percentage error (SMAPE) decreased by 48.30%, 56.64%, and 47.13%, respectively. The R² of the SSA-XGBoost model in predicting the bond strength between FRP bars and ordinary concrete, geopolymer concrete, coral concrete, and ultra-high performance concrete (UHPC) reached 0.980, 0.985, 0.956, and 0.981 respectively, indicating high prediction accuracy. In the SSA XGBoost model, the SHAP value of concrete compressive strength is 3.79. It is the parameter that contributes the most to bond strength. While the SHAP value of pull-out speed is 0.32. It is the parameter that contributes the least to bond strength. The SSA-XGBoost model efficiently and accurately predicts the bond strength between FRP bars and multiple types of concrete. And it provides interpretable analysis of the prediction model through SHAP. The research results can provide a reference for the design of bond strength between FRP bars and concrete in practical engineering.