In this paper, the effects of different influencing factors and factor interaction on the compressive strength and permeability of recycled aggregate pervious concrete (RAPC) were studied based on the response surface method (RSM). By selecting the maximum aggregate size, water cement ratio and target porosity as design variables, combined with laboratory tests and numerical analysis, the influences of three factors on the compressive strength and permeability coefficient of RAPC were revealed. The regression equation of compressive strength and permeability coefficient of recycled aggregate pervious concrete were established based on RSM, and the response surface model was optimized to determine the optimal ratio of RAPC under the conditions of meeting the mechanical and permeability properties. The results show that the mismatch item of the model is not significant, the model is credible, and the accuracy and reliability of the test are high, but the degree of uncorrelation between the test data and the model is not obvious. The sensitivity of the three factors to the compressive strength is water cement ratio > maximum coarse aggregate particle size > target porosity, and the sensitivity to the permeability coefficient is target porosity > maximum coarse aggregate particle size > water cement ratio. The absolute errors of the model prediction results and the model optimization results are 1.28 MPa and 0.19 mm/s, and the relative errors are 5.06% and 4.19%, respectively. With high accuracy, RSM can match the measured results of compressive strength and permeability coefficient of RAPC.

Recycled concrete is a heterogeneous composite material, and the composition and volume fraction of each phase affect its macroscopic properties. In this paper, ANSYS APDL was used to construct a two-dimensional numerical model of recycled aggregate concrete with different replacement rates of recycled aggregate (0%, 25%, 50%, 75% and 100%), and a uniaxial compression test was carried out to explore the relationship between recycled aggregate content and its macroscopic mechanical behavior. On this basis, the numerical simulation of different strain rates (0.1 s⁻¹, 0.05 s⁻¹, 0.01 s⁻¹, 0.005 s⁻¹ and 0.001 s⁻¹) was carried out. It was found that with the increase in the recycled aggregate replacement rate, the peak stress decreases first and then increases, and the peak strain increases continuously. When the replacement rate of recycled aggregate exceeds 50%, the overall damage area of the material increases rapidly. The strain rate will change the path of the micro-crack initiation and expansion of recycled concrete, as well as the process of damage accumulation and evolution. As a result, the unit area and shape of recycled concrete are different at different strain rates, and the damage degree of each phase material is also different.

Ultra-high-performance concrete (UHPC) relies on multivariable mix design and curing regimes, which makes empirical estimation of compressive strength increasingly unreliable when material systems vary. In this context, unlike previous studies that only applied standard eXtreme Gradient Boosting (XGBoost), this study introduces an advanced hybrid optimization strategy, Bayesian Optimization and Hyperband (BOHB), which combines the sample efficiency of Bayesian optimization with the resource allocation mechanism of Hyperband, and incorporates SHapley Additive exPlanations (SHAP) for influencing factor analysis, thereby proposing a BOHB-XGBoost framework integrated with SHAP analysis. The proposed model demonstrates excellent predictive accuracy and stability, achieving a coefficient of determination R2 of 0.967 and a Mean Absolute Percentage Error (MAPE) of 4.86% on the test set. Comparative experimental results indicate that this model shows improved and more stable performance than mainstream machine learning methods, including standard XGBOOST, Random Forest, Gradient Boosting Decision Trees, Support Vector Regression and a Multilayer Perceptron across multiple evaluation metrics. This fully demonstrates the significant advantages of the BOHB optimization framework in processing structured tabular data. By balancing exploration and exploitation during the hyperparameter search process, this strategy effectively overcomes the limitations of traditional manual tuning and significantly enhances the model’s generalization capability. Furthermore, the SHAP-based interpretability analysis identifies the key factors influencing UHPC strength, such as the water-to-cement ratio and steel fiber content. These findings are highly consistent with the experimental results conducted in this study, thereby validating the model’s physical plausibility and engineering applicability. This research not only provides an efficient and reliable method for predicting UHPC strength but also offers a transparent and interpretable scientific basis for the rational optimization of mix proportion design.