钢纤维混凝土梁抗剪强度的集成学习预测方法

狄春锋

doi:10.13204/j.gyjzG21112303

钢纤维混凝土梁抗剪强度的集成学习预测方法

doi: 10.13204/j.gyjzG21112303

狄春锋

中铁建设集团有限公司, 北京 100043

基金项目:

海南省科技厅重点研发项目(ZDYF2022SHFZ337)。

详细信息

作者简介:
狄春锋,男,1978年出生,高级工程师。电子信箱:459916706@qq.com

计量
- 文章访问数: 32
- HTML全文浏览量: 7
- PDF下载量: 2
- 被引次数: 0
出版历程
- 收稿日期: 2021-11-23

An Ensemble Learning Prediction Method for Shear Strength of Steel Fiber Reinforced Concrete Beams

DI Chunfeng

China Railway Construction Group Co., Ltd., Beijing 100043, China

摘要

摘要: 传统经验方法预测钢纤维混凝土梁的抗剪强度多基于有限的数据且难处理强非线性问题,因此难以有效快速地预测抗剪强度。为此,开发了一种基于集成学习的钢纤维混凝土梁的抗剪强度预测方法,数据包含来源于文献的330组梁。输入参数包含梁的有效尺寸、纵向配筋率、混凝土抗压强度、骨料尺寸、剪跨比、钢纤维系数及抗拉强度。首先对数据集进行热力图与相关系数分析,得到各输入参数之间几乎无冗余,均可作为有效输入参数进行建模,并得到了输入与输出参数之间的线性关系程度;然后将数据集分割为测试集和训练集,分别用于不同的集成学习模型进行计算并记录运行过程;最后将结果与传统回归方法进行比较。结果表明,集成学习中GradientBoost算法预测抗剪性能的准确度最高,达到了0.950,比传统回归方法的平均准确度要高,证明该预测方法可用于抗剪性能预测。
- 钢纤维混凝土 /
- 抗剪强度 /
- 集成学习 /
- 热力图
Abstract: Traditional empirical methods to predict the shear strength of steel fiber reinforced concrete (SFRC) beams are based on limited data and difficult to deal with strong nonlinear problems, so it is difficult to predict the shear strength effectively and quickly. A shear strength prediction method of SFRC beams based on ensemble learning was developed. The data included 330 groups of beams from literature. The input parameters included the effective size of the beam, longitudinal reinforcement ratio, concrete compressive strength, aggregate size, shear-span ratio, steel fiber factor and tensile strength of fibers. Firstly, the thermodynamic diagram and correlation coefficient of the data set were analyzed, and it was found that there was almost no redundancy between the input parameters, which could be used as effective input parameters for modeling, and the linear relations between the input and output parameters was obtained; then the data set was divided into test set and training set for different ensemble learning models to calculate and record the running process; finally, the results were compared with traditional regression methods. The results showed that the accuracy of the GradientBoost algorithm in predicting the shear strength in the ensemble learning was the highest, reaching 0.950, which was higher than the average accuracy of the traditional regression method. It proved that the method could be used to predict the shear strength of SFRC.
- steel fiber reinforced concrete /
- shear strength /
- ensemble learning /
- thermodynamic diagram

HTML全文

参考文献(18)

[1]	SHATNAWI A, ALKASSAR H M, AL-ABDALY N M, et al. Shear strength prediction of slender steel fiber reinforced concrete beams using a gradient boosting regression tree method[J]. Buildings, 2022, 12(5):550-550.
[2]	LEE S, LEE C. Prediction of shear strength of FRP-reinforced concrete flexural members without stirrups using artificial neural networks[J]. Engineering Structures, 2014, 61:99-112.
[3]	WANG Q, SONG H L, LU C L,et al. Shear performance of reinforced ultra-high performance concrete rectangular section beams[J]. Structures, 2020, 27:1184-1194.
[4]	NADERPOUR H, ALAVI S A. A proposed model to estimate shear contribution of FRP in strengthened RC beams in terms of Adaptive Neuro-Fuzzy Inference System[J]. Composite Structures, 2017, 170:215-227.
[5]	CAMPIONE G, CANNELLA F, CAVALERI L. Shear and flexural strength prediction of corroded R.C. beams[J]. Construction and Building Materials, 2017, 149:395-405.
[6]	CAMPIONE G, CANNELLA F. Engineering failure analysis of corroded R.C. beams in flexure and shear[J]. Engineering Failure Analysis, 2018, 86:100-114.
[7]	鲍跃全,李惠.人工智能时代的土木工程[J].土木工程学报, 2019, 52(5):1-11.
[8]	ÇEVIK A, KURTOĞLU A E, BILGEHAN M, et al. Support vector machines in structural engineering:a review[J]. Journal of Civil Engineering and Management, 2015, 21:261-281.
[9]	BASHIR R, ASHOUR A. Neural network modelling for shear strength of concrete members reinforced with FRP bars[J]. Composites Part B:Engineering, 2012, 43:3198-3207.
[10]	JALAL M, RAMEZANIANPOUR A A. Strength enhancement modeling of concrete cylinders confined with CFRP composites using artificial neural networks[J]. Composites Part B:Engineering, 2012, 43:2990-3000.
[11]	YASEEN Z M, TRAN M T, KIM S, et al. Shear strength prediction of steel fiber reinforced concrete beam using hybrid intelligence models:a new approach[J]. Engineering Structures, 2018, 177:244-255.
[12]	ZHAO J, NGUYEN H, NGUYEN-THOI T, et al. Improved Levenberg-Marquardt backpropagation neural network by particle swarm and whale optimization algorithms to predict the deflection of RC beams[EB/OL]. Engineering with Computers, 2021[2023-10-20]. https://link.springer.com/article/10.1007/s00366-020-01267-6.
[13]	CHEN H, DENG T, DU T, et al. An RF and LSSVM-NSGA-II method for the multi-objective optimization of high-performance concrete durability[EB/OL]. Cement and Concrete Composites, 2022[2023-10-20]. https://www.sciencedirect.com/science/article/abs/pii/S0958946522000427.
[14]	CHEN N, ZHAO S, GAO Z,et al. Virtual mix design:Prediction of compressive strength of concrete with industrial wastes using deep data augmentation[EB/OL]. Construction and Building Materials, 2022[2023-10-20]. https://www.sciencedirect.com/science/article/abs/pii/S0950061822002720.
[15]	GONG H, SUN Y, DONG Y, et al. An efficient and robust method for predicting asphalt concrete dynamic modulus[J]. International Journal of Pavement Engineering, 2022, 23(8):2565-2576.
[16]	胡旭东,张起森,范勇军.HMA动态模量Witczak和Hirsch预测模型[J].中外公路, 2003, 26(6):204-207.
[17]	HUANG H, BURTON H. Classification of in-plane failure modes for reinforced concrete frames with infills using machine learning[EB/OL]. Journal of Building Engineering, 2019[2023-10-20]. https://www.sciencedirect.com/science/article/abs/pii/S2352710218313652.
[18]	ASTERIS P G, SKENTOU A D, BARDHAN A, et al. Predicting concrete compressive strength using hybrid ensembling of surrogate machine learning models[EB/OL]. Cement and Concrete Research, 2021[2023-10-20]. https://www.sciencedirect.com/science/article/abs/pii/S0008884621000983.