地震研究

现有钢筋混凝土(RC)柱抗侧移承载力预测模型缺乏泛化性能,延性柱抗弯承载力的预测模型不能用于非延性柱的抗剪承载力,反之亦然。机器学习(ML)方法能够解决这一问题,但由于无法自动剔除冗余和不相关特征,使得ML模型复杂度高且容易过拟合。为此,提出一种樽海鞘算法优化支持向量机(SSALS-SVM)方法,基于给定的数据集,SSALS-SVM能利用樽海鞘优化算法(SSA)自动剔除冗余和不相关的特征,筛选最具代表性且各特征之间相关性弱的特征子集形成最优特征组合,同时对控制模型非线性拟合能力的超参数进行优化。优化后的模型既能识别出影响延性和非延性RC柱抗侧移承载力的设计变量,又能反映最优特征组合与抗侧移承载力间的非线性映射关系。为了验证SSALS-SVM方法的泛化性能,基于248个RC柱抗侧移承载力试验数据,分别与现有的RC柱抗侧移承载力预测模型进行对比,结果表明,SSALS-SVM比现有预测模型的泛化性能最高提升了83%。

The existing methods of predicting the lateral strength of RC columns lack generalization performance,where methods of predicting the flexural strength of ductile columns cannot be used to predict the shear strength of non-ductile columns,and vice versa.While current machine learning methods can solve this problem,they cannot automatically remove a large number of redundant and irrelevant features from the dataset,which in turn increases the complexity of the ML model and leads to overfitting.To this end,this paper proposes a new method called salp swarm algorithm-optimized least squares support vector machines(SSALS-SVM),which can remedy the aforementioned problems.Based on a given data set,SSALS-SVM can adopt the salp swarm algorithm(SSA)to automatically eliminate redundant and irrelevant features and select the most representative feature subset with weak correlation among features to form an optimal feature combination,while the hyperparameters governing the nonlinear fitting ability of LS-SVM are also optimized.In this way,the optimized prediction model can not only identify the design variables that influence the lateral strength of ductile and non-ductile columns,but also reflect the nonlinear mapping relationship between the optimal feature combination and lateral strength of RC columns.The generalization performance of proposed SSALS-SVM for predicting the lateral strength of RC columns is verified by comparing with existing prediction models based on 248 experimental data of RC columns.Numerical results show that the generalization performance of proposed SSALS-SVM can be enhanced up to 83% higher than that of existing prediction models.

引言
1 数学模型构建
2 RC柱抗侧移承载力预测
3 模型泛化性研究
4 结论

图1 超参数和输入变量组合形式<br/>Fig.1 Combination of hyperparameters and input variables

图1 超参数和输入变量组合形式
Fig.1 Combination of hyperparameters and input variables

图2 樽海鞘算法优化支持向量机模型的构建过程<br/>Fig.2 Implementation of the proposed SSALS-SVM model

图2 樽海鞘算法优化支持向量机模型的构建过程
Fig.2 Implementation of the proposed SSALS-SVM model

表1 数据集中设计变量信息<br/>Tab.1 Information of design variables

表1 数据集中设计变量信息
Tab.1 Information of design variables

图3 筛选不同柱设计变量组合时SSALS-SVM模型的RC柱抗侧移承载力预测性能分析<br/>Fig.3 The analysis of predictive performance for proposed SSALS-SVM model in the case of different combinations of selected features

图3 筛选不同柱设计变量组合时SSALS-SVM模型的RC柱抗侧移承载力预测性能分析
Fig.3 The analysis of predictive performance for proposed SSALS-SVM model in the case of different combinations of selected features

表2 不同设计变量组合对模型预测性能的影响<br/>Tab.2 Effect of different combinations of design variables on the predictive performance of the model

表2 不同设计变量组合对模型预测性能的影响
Tab.2 Effect of different combinations of design variables on the predictive performance of the model

表3 RC柱抗弯和抗剪承载力计算模型<br/>Tab.3 Equations for predicting flexural and shear strength of RC columns

表3 RC柱抗弯和抗剪承载力计算模型
Tab.3 Equations for predicting flexural and shear strength of RC columns

图4 不同预测模型的性能分析<br/>Fig.4 Performance analysis of different predictive models

图4 不同预测模型的性能分析
Fig.4 Performance analysis of different predictive models

图5 关键设计变量对SSALS-SVM模型泛化性能的影响<br/>Fig.5 Influence of key design variables on the generalization performance of proposed SSALS-SVM model

图5 关键设计变量对SSALS-SVM模型泛化性能的影响
Fig.5 Influence of key design variables on the generalization performance of proposed SSALS-SVM model

表4 各模型在不同破坏模式时的R2,RMSE,MAE值<br/>Tab.4 R2,RMSE and MAE values for each model in the case of different failure modes

表4 各模型在不同破坏模式时的R2,RMSE,MAE值
Tab.4 R2,RMSE and MAE values for each model in the case of different failure modes

邓明科,张阳玺,胡红波.2018.高延性混凝土加固钢筋混凝土柱抗剪承载力计算[J].工程力学,35(3):159-166.
Deng M K,Zhang Y X,Hu H B.2018.Experimental study and calculation of the shear capacity of RC columns strengthened with high ductile concrete[J].Engineering Mechanics,35(3):159-166.(in Chinese)
甘丹,周绪红,刘界鹏,等.2018.钢管约束钢筋混凝土柱受剪承载力计算[J].建筑结构学报,(9):96-103,111.
Gan D,Zhou X H,Liu J P,et al.2018.Calculation for shear strength of reinforced-concrete columns constrained by steel tubes[J].Journal of Building Structures,(9):96-103,111.(in Chinese)
薛亦聪,杨勇,于云龙.2020.型钢混凝土短柱受剪承载力计算模型研究[J].建筑结构学报,41(S1):162-170.
Xue Y C,Yang Y,Yu Y L.2020.Theoretical model on shear strength of steel reinforced concrete short columns[J].Journal of Building Structures,41(S1):162-170.(in Chinese)
于晓辉,王猛,宁超列.2022.基于机器学习的钢筋混凝土柱失效模式两阶段判别方法[J].建筑结构学报,43(8):220.
Yu X H,Wang M,Ning C L.2022.A machine-learning-based two-step method for failure mode classification of reinforced concrete columns[J].Journal of Building Structures,43(8):220.(in Chinese)
张勤,贡金鑫,马颖.2014.单调和反复荷载作用下弯剪破坏钢筋混凝土柱荷载-变形关系试验研究及简化模型[J].建筑结构学报,35(3):138.
Zhang Q,Gong J X,Ma Y.2014.Study on lateral load-deformation relations of flexural-shear failure columns under monotonic and cyclic loading[J].Journal of Building Structures,35(3):138.(in Chinese)
ACI 318-22,Building code requirements for structural concrete[S].
Aval S B B,Ketabdari H,Gharebaghi S A.2017.Estimating shear strength of short rectangular reinforced concrete columns using nonlinear regression and gene expression programming[J].Structures,12:13-23.
Bae S,Bayrak O.2003.Early cover spalling in high-strength concrete columns[J].Journal of Structural Engineering,129(3):314-323.
Cortes C,Vapnik V.1995.Support vector machine[J].Machine Learning,20:273-297.
Guyon I,Elisseeff A.2003.An introduction to variable and feature selection[J].Journal of Machine Learning Research,3(Mar):1157-1182.
Huang C L,Wang C J.2006.A GA-based feature selection and parameters optimizationfor support vector machines[J].Expert Systems with Applications,31(2):231-240.
Kakavand M R A,Sezen H,Taciroglu E.2021.Data-driven models for predicting the shear strength of rectangular and circular reinforced concrete columns[J].Journal of Structural Engineering,147(1):04020301.
Ketabdari H,Karimi F,Rasouli M.2020.Shear strength prediction of short circular reinforced-concrete columns using soft computing methods[J].Advances in Structural Engineering,23(14):3048-3061.
Lee J H,Son H S.2000.Failure and strength of high-strength concrete columns subjected to eccentric loads[J].ACI Structural Journal,97(1):75-85.
Lin S W,Ying K C,Chen S C,et al.2008.Particle swarm optimization for parameter determination and feature selection of support vector machines[J].Expert Systems with Applications,35(4):1817-1824.
Luo H,Paal S G.2018.Machine learning–based backbone curve model of reinforced concrete columns subjected to cyclic loading reversals[J].Journal of Computing in Civil Engineering,32(5):04018042.
Luo H,Paal S G.2021.Metaheuristic least squares support vector machine-based lateral strength modelling of reinforced concrete columns subjected to earthquake loads[J].Structures,33:748-758.
Mirjalili S,Gandomi A H,Mirjalili S Z,et al.2017.Salp swarm algorithm:A bio-inspired optimizer for engineering design problems[J].Advances in Engineering Software,114:163-191.
Noroozieh E,Mansouri A.2019.Lateral strength and ductility of reinforced concrete columns strengthened with NSM FRP rebars and FRP jacket[J].International Journal of Advanced Structural Engineering,11(2):195-209.
Ozbakkaloglu T,Saatcioglu M.2004.Rectangular stress block for high-strength concrete[J].ACI Structural Journal,101(4):475-483.
Pan Z,Li B.2012.Truss-arch model for shear strength of shear-critical reinforced concrete columns[J].Journal of Structural Engineering,139(4):548-560.
Priestley M J N,Verma R,Xiao Y.1994.Seismic shear strength of reinforced concrete columns[J].Journal of Structural Engineering,120(8):2310-2329.
Sezen H,Moehle J P.2004.Shear strength model for lightly reinforced concrete columns[J].Journal of Structural Engineering,130(11):1692-1703.
Suykens JVA,Van Gestel T,De Brabanter J.2002.Least squares support vector machines[M].Beijing:World Scientific Publishing House.

备注

引言

1 数学模型构建

2 RC柱抗侧移承载力预测

3 模型泛化性研究

4 结论

期刊信息