基于GMDH的FRP板-混凝土黏结强度预测模型

doi:10.19936/j.cnki.2096-8000.20231128.014

摘要/Abstract

摘要： 近年来,纤维增强复合材料(FRP)已被广泛应用于既有混凝土结构加固。而现有FRP与混凝土界面黏结强度模型的预测精度偏低,无法为FRP的实际应用提供有效的参考。为此,本文建立了一个由855组试验数据组成的大型数据库,并在此基础上利用数据分组处理算法(GMDH)提出了一个高精度预测模型。为验证GMDH模型是否能为FRP-混凝土界面的黏结强度计算提供更有价值的参考,本文将GMDH模型与现有七个文献模型进行对比,并采用决定系数(R²)、变异系数(COV)和相对平方根误差(RRSE)等三个统计指标对模型的预测结果进行评价。结果表明,与现有文献模型相比,本文所提GMDH模型的R²,COV和RRSE值至少改进了11.9%、30.9%和35.3%。因此,GMDH模型能够为FRP的实际应用提供更加有效的参考。

关键词: 黏结强度, FRP板, 混凝土, GMDH, 机器学习, 复合材料

Abstract: In recent years, fiber-reinforced polymer (FRP) has been widely used in concrete reinforcement. However, the prediction accuracy of the existing FRP-concrete interface bond strength prediction models is low, which cannot provide an effective reference for the practical applications of FRP composites. Therefore, a large database consisting of 855 sets of test data was established in this study, and a bond strength model with high prediction accuracy was proposed by using group method of data handling (GMDH). In order to verify whether the GMDH model can provide a more valuable reference for the calculation of the bond strength of the FRP-concrete interface, the GMDH model was compared with seven existing literature models, and coefficient of determination (R²), coefficient of variation (COV) and relative square root error (RRSE) were used to evaluate the prediction results of these models. The results showed that, compared with the existing literature models, the R², COV and RRSE values of the proposed GMDH model were improved by at least 11.9%, 30.9% and 35.3%, respectively. Therefore, the GMDH model can provide a more effective reference for the practical applications of FRP composites.

Key words: bond strength, FRP, concrete, GMDH, machine learning, composites

中图分类号:

TB332

易晓园, 张爱玲. 基于GMDH的FRP板-混凝土黏结强度预测模型[J]. 复合材料科学与工程, 2023, 0(11): 102-107.

YI Xiaoyuan, ZHANG Ailing. Prediction model of FRP-concrete bond strength based on GMDH[J]. COMPOSITES SCIENCE AND ENGINEERING, 2023, 0(11): 102-107.

参考文献

[1] 李路路. CFRP筋与混凝土粘结性能研究[D]. 哈尔滨: 哈尔滨工业大学, 2020.
[2] 高丹盈, BRAHIM B. 纤维聚合物筋与混凝土粘结性能的影响因素[J]. 工业建筑, 2001, 31(2): 9-14.
[3] ACHILLIDES Z, PILAKOUTAS K. Bond behavior of fiber reinforced polymer bars under direct pullout conditions[J]. Journal of Composites for Construction, 2004, 8(2): 173-181.
[4] SMITH S, TENG J. FRP-strengthened RC beams. Ⅰ: review of debonding strength models[J]. Engineering Structures, 2002, 24(4): 385-395.
[5] CHEN C, LI X, ZHAO D B, et al. Mechanism of surface preparation on FRP-concrete bond performance: A quantitative study[J]. Composites Part B: Engineering, 2019, 163: 193-206.
[6] ZHOU Y W, ZHENG Y W, PAN J, et al. Experimental investigations on corrosion resistance of innovative steel-FRP composite bars using X-ray microcomputed tomography[J]. Composites Part B: Engineering, 2019, 161: 272-284.
[7] ZHAO M, ANSARI F. Bond properties of FRP fabrics and concrete joints[C]//Canadian Association for Earthquake Engineering, Department of Civil Engineering. Storage Medium: Proceedings of the 13th World Conference on Earthquake Engineering. Vancouver, B.C., Canada:Canadian Association for Earthquake Engineering, 2004.
[8] 王言磊, 郝庆多, 欧进萍. 具有粗砂层的FRP板与混凝土粘结性能试验研究[J]. 武汉理工大学学报(交通科学与工程版), 2007(3): 393-396.
[9] 韦华洋. 纤维布与混凝土正拉粘结性能的试验研究[D]. 泉州: 华侨大学, 2011.
[10] 陈启壮. 冻融环境下CFRP板-ECC-混凝土复合界面粘结性能试验研究[D]. 郑州: 郑州大学, 2021.
[11] SERACINO R, SAIFULNAZ M R R, OEHLERS D J. Generic debonding resistance of EB and NSM plate-to-concrete joints[J]. Journal of Composites for Construction, 2007, 11(1): 62-70.
[12] WU Y F, JIANG C. Quantification of bond-slip relationship for externally bonded FRP to-concrete joints[J]. Journal of Composites for Construction, 2013, 17(5): 673-686.
[13] LIN J P, WU Y F, SMITH S T. Width factor for externally bonded FRP-to-concrete joints[J].Construction and Building Materials, 2017, 155: 818-829.
[14] MASHREI M A, SERACINO R, RAHMAN M S. Application of artificial neural networks to predict the bond strength of FRP-to-concrete joints[J]. Construction and Building Materials, 2013, 40(3): 812-821.
[15] GOLAFSHANI E M, RAHAI A, SEBT M H. Artificial neural network and genetic programming for predicting the bond strength of GFRP bars in concrete[J]. Materials and Structures, 2015, 48(5): 1581-1602.
[16] 段伟, 蔡国军, 袁俊, 等. 基于GMDH的地震液化场地侧向变形预测模型[J]. 东南大学学报(自然科学版), 2021, 51(2): 306-311.
[17] IVAKHNENKO A G, SAVCHENKO E A. Problems of further GMDH algorithms development[J]. System Analysis Modelling Simulation, 2003, 43(10): 1301-1309.
[18] 陈建平, 杨宜民, 张会章, 等. 一种基于GMDH模型的神经网络学习算法[J]. 云南大学学报(自然科学版), 2008(6): 569-574.
[19] 李少远, 王景成. 智能控制[M]. 北京: 机械工业出版社, 2005.
[20] ZHOU Y W, ZHENG S B, HUANG Z Y, et al. Explicit neural network model for predicting FRP-concrete interfacial bond strength based on a large database[J]. Composite Structures, 2020, 240:111998.
[21] MAEDA T, ASANO Y, SATO Y, et al. A study on bond mechanism of carbon fiber sheet[C]//Japan Concrete Institute, Department of Engineering. Materials and Structures: 3rd International Symposium on Non-Metallic (FRP) Reinforcement for Concrete Structures. Japan: Japan Concrete Institute, 1997: 279-285.
[22] 杨勇新, 岳瑞, 胡云昌. 碳纤维布与混凝土粘结性能的试验研究[J]. 建筑结构学报, 2001, 22(3): 36-42.
[23] CHEN J F, TENG J G. Anchorage strength models for FRP and steel plates bonded to concrete[J]. Journal of Structural Engineering, 2001, 127(7): 784-791.
[24] 陆新征. FRP-混凝土界面行为研究[D]. 北京: 清华大学, 2005.
[25] DAI J, UEDA T, SATO Y. Development of the nonlinear bond stress-slip model of fiber reinforced plastics sheet-concrete interfaces with a simple method[J]. Journal of Composites for Construction, 2005, 9(1): 52-62.
[26] WU Z, ISLAM S M, SAID H. A three-parameter bond strength model for FRP-concrete interface[J]. Journal of Reinforced Plastics and Composites, 2009, 28(19): 2309-2323.