FRP加固RC梁抗剪承载力智能预测方法

An Intelligent prediction method for shear bearing capacity of FRP reinforced RC beams

纤维增强复合材料(Fiber Reinforced Polymer, FRP)加固钢筋混凝土(Reinforced Concrete, RC)梁抗剪承载力的计算存在公式复杂、计算流程繁琐等问题。为了解决这些问题,文章采用集成学习方法建立了FRP加固RC梁抗剪承载力的预测模型,通过相关文献收集试验数据并进行预处理,利用RF、GBDT、XGBoost 3种算法预测,同时应用贝叶斯优化算法优化模型超参数并对比不同模型预测效果。结果表明:与GBDT、XGBoost模型相比,RF模型具有更好的预测效果;贝叶斯优化算法能够显著提升模型的预测性能,而优化后的RF模型具有更高的预测精度与泛化能力。研究成果可为FRP加固RC梁的研究和应用提供参考,同时也为承载力的预测提供了新的途径。

The calculation of the shear bearing capacity of Fiber Reinforced Polymer(FRP)reinforced Reinforced Concrete(RC)beams have problems of complex formulas and tedious calculation processes.To solve these problems,ensemble learning methods were used to develop prediction models of the shear bearing capacity of Fiber Reinforced Polymer(FRP)strengthened RC beams.Experimental data were collected from relevant literature and preprocessed.RF,GBDT,and XGBoost algorithms were utilized for prediction and Bayesian optimization algorithm was used to optimize hyperparameters and compare the prediction results of different models.Results show that RF model has better prediction performance than GBDT and XGBoost models.Bayesian optimization algorithm can significantly improve the prediction performance of the models.The optimized RF model has higher prediction accuracy and generalization ability,which can provide references for research and application of FRP reinforced RC beams,offer new approaches for capacity prediction.