基于<strong>Stacking</strong>集成模型融合的钢筋混凝土锈胀开裂预测方法

doi:10.11902/1005.4537.2024.017

中国腐蚀与防护学报

2024, Vol. 44

Issue (6): 1601-1609 CSTR: 32134.14.1005.4537.2024.017 DOI: 10.11902/1005.4537.2024.017

研究报告

本期目录 | 过刊浏览

基于Stacking集成模型融合的钢筋混凝土锈胀开裂预测方法

梁梓豪¹^,²^,³, 应宗权¹^,²^,³, 刘梅梅¹^,²^,³(

), 杨帅¹^,²^,³

1.中交四航工程研究院有限公司广州 510230
2.中交集团建筑材料重点实验室广州 510230
3.水工构造物耐久性技术交通运输行业重点实验室广州 510230

Prediction Method for Reinforced Concrete Corrosion-induced Crack Based on Stacking Integrated Model Fusion

LIANG Zihao¹^,²^,³, YING Zongquan¹^,²^,³, LIU Meimei¹^,²^,³(

), YANG Shuai¹^,²^,³

1. CCCC Fourth Harbor Engineering Institute Co., Ltd., Guangzhou 510230, China
2. Key Laboratory of Construction Materials, CCCC, Guangzhou 510230, China
3. Key Laboratory of Harbor & Marine Structure Durability Technology, Ministry of Transport, Guangzhou 510230, China

引用本文:

梁梓豪, 应宗权, 刘梅梅, 杨帅. 基于Stacking集成模型融合的钢筋混凝土锈胀开裂预测方法[J]. 中国腐蚀与防护学报, 2024, 44(6): 1601-1609.
Zihao LIANG, Zongquan YING, Meimei LIU, Shuai YANG. Prediction Method for Reinforced Concrete Corrosion-induced Crack Based on Stacking Integrated Model Fusion[J]. Journal of Chinese Society for Corrosion and protection, 2024, 44(6): 1601-1609.

全文: PDF(3787 KB) HTML

摘要:

为了解决传统经验公式模型对钢筋混凝土锈胀开裂预测方法存在的公式不统一、精度有限等问题的局限性，本文提出一种基于Stacking集成模型融合的钢筋混凝土锈胀开裂预测方法。首先，通过文献收集的223组钢筋混凝土锈胀开裂试验数据进行数据预处理；其次，基于支持向量回归(SVR)、随机森林(RF)和极端梯度提升树(XGBoost)算法分别进行贝叶斯优化超参数、模型训练及评估，采用决定系数 $R 2$ 、平均绝对误差(MAE)和均方根误差(RMSE)对比分析3种机器学习模型的预测性能，并在此基础上搭建基于Stacking融合多种算法的预测模型；最后，对所提出的机器学习模型和传统经验公式模型进行泛化性能验证，并探讨基于XGBoost模型的可解释性分析。结果表明：与其他机器学习模型相比，基于Stacking集成模型的预测精度和泛化性能较好，且可解释性分析结果说明模型预测逻辑与实际工程经验较为吻合。研究结果有助于提高钢筋混凝土锈胀开裂预测精度，可为决策者在实际工程提供科学的理论指导。

关键词 ：钢筋混凝土, 锈胀开裂, 机器学习, Stacking算法, XGBoost算法, 可解释性分析

Abstract：

In predicting corrosion-induced cracking of reinforced concrete, traditional empirical formulas used are varied with limited precision of prediction. To address these limitations, this paper presents a method based on the stacking of models to predict the cracking of reinforced concrete due to corrosion induced expansion. Firstly, 223 sets of test data on the cracking of reinforced concrete due to corrosion induced expansion were collected from published articles and processed in advance. Next, Bayesian optimization of hyperparameters, model training, and evaluation were conducted separately based on Support Vector Regression (SVR), Random Forest (RF), and Extreme Gradient Boosting (XGBoost) algorithms. Determination coefficient (R²), mean absolute error (MAE), and root mean square error (RMSE) were utilized for a comparative analysis of the prediction performances of three machine learning models. On this basis, a prediction model integrating multiple algorithms with the Stacking method was proposed. Finally, the generalization performances of the proposed prediction model and traditional empirical formula models were verified, and the XGBoost model was employed to analyze the interpretability of the proposed model. As revealed in the results, the proposed model has better prediction accuracy and generalization performance than other machine learning models. The interpretability analysis result demonstrates that the prediction of the proposed model logic matches the practical engineering experience. This finding is conducive to improve the prediction accuracy of thecorrosion-induced cracking of reinforced concrete, and can provide scientific theoretical guidance for decision-makers in practical engineering.

Key words： reinforced concrete corrosion-induced cracking machine learing stacking algorithm XGBoost algorithm interpretability analysis

收稿日期: 2024-01-12 32134.14.1005.4537.2024.017

ZTFLH:

TU375

基金资助:国家重点研发计划(2022YFB2603000)

通讯作者: 刘梅梅，E-mail: lmeimei@cccc4.com，研究方向为钢筋混凝土结构耐久性

Corresponding author: LIU Meimei, E-mail: lmeimei@cccc4.com

作者简介: 梁梓豪，男，1997年生，硕士，助理工程师

	服务

	把本文推荐给朋友
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	梁梓豪
	应宗权
	刘梅梅
	杨帅
44.8K

链接本文:

https://www.jcscp.org/CN/10.11902/1005.4537.2024.017 或 https://www.jcscp.org/CN/Y2024/V44/I6/1601

图1 模型输入和输出特征数据统计

图2 相关性分析热力图

图3 Stacking集成模型框架

Arithmetic	Hyperparameterization	Optimal parameter values	$R 2$	RMSE/%	MAE/%
Stacking	/	/	0.906	1.17	0.88
XGBoost	n_estimator	222	0.883	1.31	0.95
	max_depth	9
	gamma	3
	learning_rate	0.3
RF	max_depth	9	0.837	1.55	1.09
	min_samples_split	2
	min_sample_leaf	1
	n_estimators	125
SVR	C	32	0.773	1.83	1.46
	Gamma	1	0.773	1.83	1.46

表1 机器学习模型预测结果比较

图4 标准化泰勒图

Traditional empirical formulas	Calculation formula
Model one^[5]	$β = 1 d 32.43 + 0.303 f c u + 0.65 c + 27.45 w$
Model two^[27]	$β = 94.82 × 1 . 018 c f c u 0.248 d - 1.588$
Model three^[38]	$w = 4.5045 δ + 0.04955, β = 1 - 1 - 2 δ d 2 × 100 %$

表2 3种传统经验公式

图5 机器学习与传统经验公式模型的预测值、真实值和绝对误差

表3 机器学习与经验公式模型误差比较

图6 模型泛化性能验证比较

图7 单特征依赖图

1	Bai Y L, Jin W L, Yu Y F, et al. Reliability analysis of damaged concrete beams based on the rust crack width [J]. J. Civ. Environ. Eng., 2023: 1-8, doi: 10.11835/j.issn.2096-6717.2023.076
1	(柏玉良, 金伟良, 余一凡等. 基于锈胀裂缝宽度的损伤混凝土梁可靠度分析 [J]. 土木与环境工程学报(中英文), 2023: 1-8, doi: 10.11835/j.issn.2096-6717.2023.076
2	Jin W L, Zhao Y X. Durability of Concrete Structures [M]. 2nd ed. Beijing: Science Press, 2014: 1
2	(金伟良, 赵羽习. 混凝土结构耐久性 [M]. 2版. 北京: 科学出版社, 2014: 1)
3	Zhang Y, Jiang L X. A practical mathematical model of concrete carbonation depth based on the mechanism [J]. Ind. Constr., 1998, 28: 16
3	(张誉, 蒋利学. 基于碳化机理的混凝土碳化深度实用数学模型 [J]. 工业建筑, 1998, 28: 16)
4	Parrott L J. A study of carbonation-induced corrosion [J]. Mag. Concrete Res., 1994, 46: 23
5	Hui Y L. Assessment and predicted experimental study on corrosive degree of reinforcements in concrete structures [J]. Ind. Constr., 1997, 27(6): 6
5	(惠云玲. 混凝土结构中钢筋锈蚀程度评估和预测试验研究 [J]. 工业建筑, 1997, 27(6): 6)
6	Di X T, Zhou Y. Inspection and Maintenance of Old Buildings [M]. Beijing: Seismological Press, 1992: 78
6	(邸小坛, 周燕. 旧建筑物的检测加固与维护 [M]. 北京: 地震出版社, 1992: 78)
7	Wang S. Research and durability assessment of rust expansion cracks in reinforced concrete structures [D]. Shanghai: Tongji University, 2000
7	(王深. 钢筋混凝土结构锈胀裂缝的研究及耐久性评估 [D]. 上海: 同济大学, 2000)
8	Tong L X. Research on the damage of concrete and the influence for the longitudinal reinforcement corrosion by stirrup corrosion [D]. Xi’an: Xi’an University of Architecture and Technology, 2013
8	(同立向. 箍筋锈蚀引起的混凝土保护层损伤及对纵筋锈蚀的影响研究 [D]. 西安: 西安建筑科技大学, 2013)
9	Zhang W G, Tang L B, Chen F Y, et al. Prediction for TBM penetration rate using four hyperparameter optimization methods and random forest model [J]. J. Basic Sci. Eng., 2021, 29: 1186
9	(仉文岗, 唐理斌, 陈福勇等. 基于4种超参数优化算法及随机森林模型预测TBM掘进速度 [J]. 应用基础与工程科学学报, 2021, 29: 1186)
10	Wang S S, Ma S J, Che K, et al. Application status of machine learning in field of natural environment corrosion assessment and prediction [J]. J. Chin. Soc. Corros. Prot., 2023, 43: 441
10	(王莎莎, 马帅杰, 车琨等. 机器学习在自然环境腐蚀评估与预测领域的应用现状 [J]. 中国腐蚀与防护学报, 2023, 43: 441) doi: 10.11902/1005.4537.2022.147
11	Yao Y, Liu G J, Li S Z, et al. Research progress on corrosion prediction model of metallic materials for electrical equipment [J]. J. Chin. Soc. Corros. Prot., 2023, 43: 983
11	(姚勇, 刘国军, 黎石竹等. 金属材料腐蚀预测模型研究进展 [J]. 中国腐蚀与防护学报, 2023, 43: 983)
12	Liu X, Wang S M, Lu L, et al. Development on machine learning for durability prediction of concrete materials [J]. J. Chin. Ceram. Soc., 2023, 51: 2062
12	(刘晓, 王思迈, 卢磊等. 机器学习预测混凝土材料耐久性的研究进展 [J]. 硅酸盐学报, 2023, 51: 2062)
13	Liu Y, Zou X X, Yang Z W, et al. Machine learning embedded with materials domain knowledge [J]. J. Chin. Ceram. Soc., 2022, 50: 863
13	(刘悦, 邹欣欣, 杨正伟等. 材料领域知识嵌入的机器学习 [J]. 硅酸盐学报, 2022, 50: 863)
14	Xu X H, Hu Z L, Liu J P, et al. Concrete strength prediction of the three gorges dam based on machine learning regression model [J]. Mater. Rep., 2023, 37(2): 45
14	(徐潇航, 胡张莉, 刘加平等. 基于机器学习回归模型的三峡大坝混凝土强度预测 [J]. 材料导报, 2023, 37(2): 45)
15	Bai T, Luo X B, Xing G H. Prediction of abrasion resistance of pervious concrete based on machine learning [J]. Bull. Chin. Ceram. Soc., 2024, 43: 138
15	(白涛, 罗小宝, 邢国华. 基于机器学习的透水混凝土耐磨性能预测 [J]. 硅酸盐通报, 2024, 43: 138)
16	Liu K H, Dai Z H, Zhang R B, et al. Prediction of the sulfate resistance for recycled aggregate concrete based on ensemble learning algorithms [J]. Constr. Build. Mater., 2022, 317: 125917
17	Li Z J, Qi J N, Hu Y Q, et al. Estimation of bond strength between UHPC and reinforcing bars using machine learning approaches [J]. Eng. Struct., 2022, 262: 114311
18	Huang T, Liu T B, Ai Y, et al. Modelling the interface bond strength of corroded reinforced concrete using hybrid machine learning algorithms [J]. J. Build. Eng., 2023, 74: 106862
19	Hu Y C, Liang M, Xie C R, et al. Strength prediction method of high performance concrete based on stacking model fusion [J]. Bull. Chin. Ceram. Soc., 2023, 42: 3914
19	(胡以婵, 梁铭, 谢灿荣等. 基于Stacking模型融合的高性能混凝土强度预测方法 [J]. 硅酸盐通报, 2023, 42: 3914)
20	Zhong X P, Peng L G, Yuan C B, et al. Experimental research on compressive strength of concrete damaged by coupling of chlorine-corrosion [J]. Ind. Constr., 2020, 50(12): 69
20	(钟小平, 彭蓝鸽, 袁承斌等. 氯盐-锈蚀耦合损伤混凝土抗压强度试验 [J]. 工业建筑, 2020, 50(12): 69)
21	Xiao D, Zhao T Y, Di Z Y. Effect of protective layer on crack of corroded reinforced concrete beam [J]. J. Univ. Sci. Technol. Liaoning, 2015, 38: 306
21	(肖丹, 赵天宇, 邸振禹. 保护层对锈蚀钢筋混凝土梁裂缝的影响 [J]. 辽宁科技大学学报, 2015, 38: 306)
22	Yang X M, Yang Z B, Yang L. Relationship between corrosion rate and width of corrosion cracks in corner of concrete components with different depth of cover [J]. Bull. Chin. Ceram. Soc., 2019, 38: 3332
22	(杨晓明, 杨治邦, 杨亮. 基于不同保护层厚度下混凝土构件角区锈胀裂缝宽度与钢筋锈蚀率的关系研究 [J]. 硅酸盐通报, 2019, 38: 3332)
23	Yang X M, Yang Z B, Yang L. Study on the relationship between total width of corrosion cracks in the corner of corroded reinforcement corrosion components and corrosion rate [J]. Bull. Chin. Ceram. Soc., 2018, 37: 3807
23	(杨晓明, 杨治邦, 杨亮. 锈蚀钢筋混凝土构件角部锈胀裂缝总宽度与钢筋锈蚀率之间的关系研究 [J]. 硅酸盐通报, 2018, 37: 3807)
24	Wu W. Research on concrete reinforcement corrosion process and test [D]. Wuhan: Hubei University of Technology, 2012
24	(吴伟. 混凝土钢筋锈蚀过程及试验研究 [D]. 武汉: 湖北工业大学, 2012)
25	Li Y. Study on time-varying properties of load bearing-corroded reinforced concrete structures [D]. Yangzhou: Yangzhou University, 2018
25	(李叶. 持载锈蚀钢筋混凝土结构时变性能研究 [D]. 扬州: 扬州大学, 2018)
26	Peng J X, Hu S W, Zhang J R, et al. Experimental study and prediction model of corrosion-induced crack width in RC structure [J]. J. Exp. Mech., 2014, 29: 33
26	(彭建新, 胡守旺, 张建仁等. 钢筋混凝土结构锈胀开裂宽度的试验研究及预测模型 [J]. 实验力学, 2014, 29: 33)
27	Li H B, Yang F, Zhao Y X, et al. Model of corroded expansion force at cracking on reinforced concrete structures [J]. J. Zhejiang Univ. (Eng. Sci.), 2000, 34(4): 67
27	(李海波, 鄢飞, 赵羽习等. 钢筋混凝土结构开裂时刻的钢筋锈胀力模型 [J]. 浙江大学学报(工学版), 2000, 34(4): 67)
28	Yuan L Q. The investigation and analysis of corroded and cracked reinforced concrete [D]. Xi’an: Xi’an University of Architecture and Technology, 2007
28	(袁立群. 混凝土结构锈胀裂缝分析与钢筋锈蚀预测研究 [D]. 西安: 西安建筑科技大学, 2007)
29	Patro S G K, Sahu K K. Normalization: A preprocessing stage [Z]. arXiv preprint arXiv: 1503.06462, 2015
30	Benesty J, Chen J D, Huang Y, et al. Pearson correlation coefficient [A]. CohenI, HuangYT, ChenJD, et al. Noise Reduction in Speech Processing [M]. Berlin: Springer, 2009: 1
31	Cortes C, Vapnik V. Support-vector networks [J]. Mach. Learn., 1995, 20: 273
32	Breiman L. Random forests [J]. Mach. Learn., 2001, 45: 5
33	Chen T, He T, Benesty M, et al. Xgboost: extreme gradient boosting [J]. R Package Version 0.4-2, 2015, 1: 1
34	Wolpert D H. Stacked generalization [J]. Neur. Netw., 1992, 5: 241
35	Lerman P M. Fitting segmented regression models by grid search [J]. J. Roy. Stat. Soc. Ser. C.: Appl. Stat., 1980, 29: 77
36	Bergstra J, Bengio Y. Random search for hyper-parameter optimization [J]. J. Mach. Learn. Res., 2012, 13: 281
37	Shahriari B, Swersky K, Wang Z Y, et al. Taking the human out of the loop: A review of Bayesian optimization [J]. Proc. IEEE, 2016, 104: 148
38	Andrade C, Alonso C, Molina F J. Cover cracking as a function of bar corrosion: part I-experimental test [J]. Mater. Struct., 1993, 26: 453
39	Wan S W, Xu J. Experimental study on relationship between crack width and corrosion extent of concrete members [J]. Rail. Eng., 2016, (10): 128
39	(万胜武, 徐杰. 混凝土构件锈胀裂缝宽度与锈蚀量关系的试验研究 [J]. 铁道建筑, 2016, (10): 128)

[1]	沈坚, 吴柯娴, 何晓宇, 方兴龙. 我国不同地区钢材大气腐蚀预测算法评估与筛选[J]. 中国腐蚀与防护学报, 2024, 44(4): 939-948.
[2]	商百慧, 马元泰, 孟美江, 李瑛, 娄明, 白晶. HRB400钢筋在模拟混凝土孔隙液环境中的阳极极化特征[J]. 中国腐蚀与防护学报, 2024, 44(2): 422-428.
[3]	王莎莎, 马帅杰, 车琨, 杜艳霞. 机器学习在自然环境腐蚀评估与预测领域的应用现状[J]. 中国腐蚀与防护学报, 2023, 43(3): 441-451.
[4]	丁清苗, 高宇宁, 侯文亮, 秦永祥. Cl^-浓度对钢筋混凝土在土壤中腐蚀行为的影响[J]. 中国腐蚀与防护学报, 2021, 41(5): 705-711.
[5]	程旭东,孙连方,曹志烽,朱兴吉. 钢筋非均匀锈蚀导致的混凝土保护层锈胀开裂过程分析[J]. 中国腐蚀与防护学报, 2015, 35(3): 257-264.
[6]	程旭东, 孙连方, 曹志烽, 朱兴吉, 赵立新. 沿海钢筋混凝土结构Cl^-侵蚀数值模拟方法研究[J]. 中国腐蚀与防护学报, 2015, 35(2): 144-150.
[7]	徐晶,姚武. 恒流脉冲技术检测混凝土中钢筋的腐蚀[J]. 中国腐蚀与防护学报, 2010, 30(3): 181-186.
[8]	鲁照玲 . 酸性气氛下钢筋混凝土结构腐蚀行为及其机理[J]. 中国腐蚀与防护学报, 2007, 27(2): 119-123 .
[9]	丁元力; 董泽华; 周华林 . 基于护环技术的混凝土中钢筋腐蚀监测研究[J]. 中国腐蚀与防护学报, 2006, 26(5): 257-262 .
[10]	王胜先; 林薇薇; 张鉴清 . 硫脲-二乙烯三胺缩聚物对混凝土中钢筋的缓蚀作用[J]. 中国腐蚀与防护学报, 2000, 20(1): 15-21 .

Viewed

Full text

Abstract

Cited

Shared

Discussed