基于机器学习的防护与修复混凝土寿命智能预测方法

doi:10.16552/j.cnki.issn1001-1625.2025.0777

摘要/Abstract

摘要： 为定量评估防护与修复措施对混凝土结构寿命的影响,本研究提出了一种基于机器学习的智能预测方法。基于工程先验知识的数值求解方法,构建了综合考虑环境因素、混凝土材料参数及防护与修复措施参数的10万组样本数据集,系统性评估了12种主流机器学习模型,并选用基于贝叶斯超参数优化的极端梯度提升(XGBoost)算法进行回归预测,完成了SHapley Additive exPlanations(SHAP)框架对优化后模型的深度可解释性分析。结果表明,优化后的XGBoost模型在测试集上决定系数R²可达到0.986 5。SHAP分析结果量化了各特征对寿命预测的贡献度,识别出了环境氯盐浓度、粉煤灰掺量和水胶比是影响寿命预测的关键因素,揭示了多因素间的复杂非线性关系。本研究建立的高精度、可解释机器学习预测模型,为防护与修复混凝土结构的耐久性设计、性能评估与维养决策提供了高效可靠的数据驱动分析工具。

关键词: 机器学习, 回归预测, 防护与修复, 混凝土寿命, XGBoost, SHAP

Abstract: To quantitatively evaluate the effects of protection and repair measures on the service life of concrete structures, this study proposed an intelligent prediction method based on machine learning. Based on the numerical solution method incorporating engineering prior knowledge, a dataset comprising 100 000 sample groups comprehensively considering environmental factors, concrete materials parameters, and parameters of protection and repair measures was constructed. Twelve mainstream machine learning models were systematically assessed, and the extreme gradient Boosting (XGBoost) algorithm based on Bayesian hyperparameter optimization was selected for regression prediction. Concurrently, an in-depth interpretability analysis of the optimized model was conducted using the SHapley Additive exPlanations (SHAP) framework. The results indicate that the of the optimized XGBoost model determination coefficient R² is 0.986 5 on the test set. The SHAP analysis quantify the contribution of each feature to the lifespan prediction, identifying environmental chloride concentration, fly ash content, and water-binder ratio as key factors influencing lifespan prediction. Furthermore, SHAP reveals the complex non-linear relationships among various factors. The high-precision, interpretable machine learning predicted model developed in this study provides an efficient and reliable data-driven analytical tool for the durability design, performance assessment, and maintenance decision-making for protected and repaired concrete structures.

Key words: machine learning, regression prediction, protection and repair, concrete lifespan, XGBoost, SHAP

中图分类号:

TU528

王凤娟, 李映泽, 王赟程, 石锦炎, 刘志勇, 蒋金洋. 基于机器学习的防护与修复混凝土寿命智能预测方法[J]. 硅酸盐通报, 2025, 44(11): 4235-4251.

WANG Fengjuan, LI Yingze, WANG Yuncheng, SHI Jinyan, LIU Zhiyong, JIANG Jinyang. Intelligent Prediction Method for Lifespan of Protected and Repaired Concrete Based on Machine Learning[J]. BULLETIN OF THE CHINESE CERAMIC SOCIETY, 2025, 44(11): 4235-4251.

参考文献

[1] LI J Q, WU Z M, SHI C J, et al. Durability of ultra-high performance concrete: a review[J]. Construction and Building Materials, 2020, 255: 119296.
[2] TANG S W, YAO Y, ANDRADE C, et al. Recent durability studies on concrete structure[J]. Cement and Concrete Research, 2015, 78: 143-154.
[3] CAO Y, GEHLEN C, ANGST U, et al. Critical chloride content in reinforced concrete: an updated review considering chinese experience[J]. Cement and Concrete Research, 2019, 117: 58-68.
[4] BOLZONI F, BRENNA A, ORMELLESE M. Recent advances in the use of inhibitors to prevent chloride-induced corrosion in reinforced concrete[J]. Cement and Concrete Research, 2022, 154: 106719.
[5] 刘镇. 海洋环境下混凝土氯离子渗透性评价[J]. 硅酸盐学报, 2023, 51 (11): 2846-2856.
LIU Z. Evaluation of chloride ion permeability of concrete in marine environments[J]. Journal of the Chinese Ceramic Society, 2023, 51 (11): 2846-2856 (in Chinese).
[6] ALEXANDER M, BEUSHAUSEN H. Durability, service life prediction, and modelling for reinforced concrete structures-review and critique[J]. Cement and Concrete Research, 2019, 122: 17-29.
[7] SHAKOURI M, TREJO D. A time-variant model of surface chloride build-up for improved service life predictions[J]. Cement and Concrete Composites, 2017, 84: 99-110.
[8] MA J W, YANG Q W, WANG X H, et al. Review of prediction models for chloride ion concentration in concrete structures[J]. Buildings, 2025, 15(1): 149.
[9] 杨策, 张金喜, 丁勇杰, 等. 混凝土氯离子侵蚀与冻融循环劣化的数值模拟研究综述[J]. 北京工业大学学报, 2025: 1-24.
YANG C, ZHANG J X,DING Y J, et al. Review of numerical simulation on chloride penetration and freeze-thaw cycle deterioration of concrete[J]. Journal of Beijing University of Technology, 2025: 1-24 (in Chinese).
[10] 蒋金洋, 刘志勇, 许文祥, 等. 混凝土多目标性能预测与智能设计系统[M]. 南京: 东南大学出版社, 2023.
JIANG J Y, LIU Z Y, XU W X, et al. Multi-performance prediction and intelligent design system of concrete[M]. Nanjing: Southeast University Press, 2023 (in Chinese).
[11] 张大利, 宁作君, 郑秀梅, 等. 复杂环境下钢筋混凝土结构寿命预测研究进展[J]. 混凝土, 2023(8): 8-13+22.
ZHANG D L, NING Z J, ZHENG X M, et al. Research progress on life prediction of reinforced concrete structures in complex environment[J]. Concrete, 2023(8): 8-13+22 (in Chinese).
[12] CAI R, HAN T H, LIAO W Y, et al. Prediction of surface chloride concentration of marine concrete using ensemble machine learning[J]. Cement and Concrete Research, 2020, 136: 106164.
[13] 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 (8): 2062-2073.
LIU X, WANG S M, LU L, et al. Development on machine learning for durability prediction of concrete materials [J]. Journal of the Chinese Ceramic Society, 2023, 51 (8): 2062-2073 (in Chinese).
[14] 罗大明, 李凡, 牛荻涛. 人工智能时代混凝土结构耐久性诊断研究进展[J]. 建筑结构学报, 2024, 45 (2): 1-13.
LUO D M, LI F, NIU D T. Research progress on durability diagnosis of concrete structures based on artificial intelligence[J]. Journal of Building Structures, 2024, 45 (2): 1-13 (in Chinese).
[15] 覃源, 薛存, 李遥, 等. 盐冻耦合作用下水工混凝土耐久性及寿命预测[J]. 水力发电学报, 2024, 43(2): 110-122.
QIN Y, XUE C, LI Y, et al. Durability and lifespan predictions of hydraulic concrete under salt freezing coupling effect[J]. Journal of Hydroelectric Engineering, 2024, 43(2): 110-122 (in Chinese).
[16] 宋庆功, 常斌斌, 董珊珊, 等. 机器学习及其在材料研发中的作用[J]. 材料导报, 2022, 36 (1): 183-189.
SONG Q G, CHANG B B, DONG S S, et al. Machine learning and its influence on materials research and development[J]. Materials Reports, 2022, 36 (1): 183-189 (in Chinese).
[17] LI Z Z, YOON J, ZHANG R, et al. Machine learning in concrete science: applications, challenges, and best practices[J]. NPJ Computational Materials, 2022, 8: 127.
[18] ZHANG M L, KANG R. Machine learning methods for predicting the durability of concrete materials: a review[J]. Advances in Cement Research, 2025: 1-16.
[19] TAFFESE W Z, ESPINOSA-LEAL L. Prediction of chloride resistance level of concrete using machine learning for durability and service life assessment of building structures[J]. Journal of Building Engineering, 2022, 60: 105146.
[20] TAFFESE W Z, SISTONEN E. Machine learning for durability and service-life assessment of reinforced concrete structures: recent advances and future directions[J]. Automation in Construction, 2017, 77: 1-14.
[21] ZHANG H P, LI X C, AMIN M N, et al. Analyzing chloride diffusion for durability predictions of concrete using contemporary machine learning strategies[J]. Materials Today Communications, 2024, 38: 108543.
[22] TAFFESE W Z, HILLOULIN B, VILLAGRAN ZACCARDI Y, et al. Machine learning in concrete durability: challenges and pathways identified by RILEM TC 315-DCS towards enhanced predictive models[J]. Materials and Structures, 2025, 58(4): 145.
[23] YU X R, LI J H, YU Y, et al. Advancing service life estimation of reinforced concrete considering the coupling effects of multiple factors: hybridized physical testing and machine learning approach[J]. Journal of Building Engineering, 2024, 84: 108476.
[24] AGHAEE K, ROSHAN A. Predicting time to cracking of concrete composites under restrained shrinkage: a review with insights from statistical analysis and ensemble machine learning approaches[J]. Journal of Building Engineering, 2024, 97: 110856.
[25] WANG Z Y, LIU H H, AMIN M N, et al. Optimizing machine learning techniques and SHapley Additive exPlanations (SHAP) analysis for the compressive property of self-compacting concrete[J]. Materials Today Communications, 2024, 39: 108804.
[26] 周双喜, 盛伟, 何顺爱. 基于深度学习的氯盐环境下高性能混凝土氯离子扩散系数的预测[J]. 混凝土, 2019 (7): 27-31.
ZHOU S X, SHENG W, HE S A. Prediction of chloride diffusion coefficient of high performance concrete based on depth learning in chloride environment[J]. Concrete, 2019 (7): 27-31 (in Chinese).
[27] 孔祥清, 张明亮, 康然, 等. 机器学习用于预测混凝土性能的研究进展[J]. 科学技术与工程, 2025, 25(5): 1764-1773.
KONG X Q, ZHANG M L, KANG R, et al. Review on machine learning for predicting concrete properties[J]. Science Technology and Engineering, 2025, 25(5): 1764-1773 (in Chinese).
[28] ZIOLKOWSKI P, NIEDOSTATKIEWICZ M. Machine learning techniques in concrete mix design[J]. Materials, 2019, 12(8): 1256.
[29] CHEN H Y, CAO Y, LIU Y, et al. Enhancing the durability of concrete in severely cold regions: mix proportion optimization based on machine learning[J]. Construction and Building Materials, 2023, 371: 130644.
[30] YU X R. Developing an artificial neural network model to predict the durability of the RC beam by machine learning approaches[J]. Case Studies in Construction Materials, 2022, 17: e01382.
[31] WANG Y C, WANG L X, MIAO Y C, et al. Numerical simulation of convection-diffusion coupling transport of water and chloride in coated concrete[J]. Journal of Materials in Civil Engineering, 2024, 36(12): 04024424.
[32] FENG T T, YU H F, TAN Y S, et al. Service life design for concrete engineering in marine environments of northern china based on a modified theoretical model of chloride diffusion and large datasets of ocean parameters[J]. Engineering, 2022, 17: 123-139.
[33] 中国工程建设标准化协会. 严酷环境混凝土结构耐久性设计标准: T/CECS 1203—2022[S]. 北京: 中国建筑工业出版社, 2022.
Chinese Association for Engineering Construction Standardization. Standard for design of concrete structure durability in severe environments: T/CECS 1203—2022[S]. Beijing: China Planning Press, 2022 (in Chinese).
[34] 中华人民共和国交通运输部. 公路工程混凝土结构耐久性设计规范: JTG/T 3310—2019[S]. 北京: 人民交通出版社, 2019.
Ministry of Transport of the People's Republic of China. Code for durability design of concrete structures in highway engineering: JTG/T 3310—2019[S]. Beijing: China Communications Press Co., Ltd, 2019 (in Chinese).
[35] 中华人民共和国住房和城乡建设部, 中华人民共和国国家质量监督检验检疫总局. 混凝土外加剂应用技术规范: GB 50119—2013[S]. 北京: 中国建筑工业出版社, 2013.
Ministry of Housing and Urban-Rural Development of the People's Republic of China, State Administration for Market Regulation. Code for concrete admixture application: GB 50119—2013[S]. Beijing: China Architecture & Building Press, 2013 (in Chinese).
[36] NGUYEN H, VU T, VO T P, et al. Efficient machine learning models for prediction of concrete strengths[J]. Construction and Building Materials, 2021, 266: 120950.
[37] PANDEY D, NIWARIA K, CHOURASIA B. Machine learning algorithms: a review[J]. International Research Journal of Engineering and Technology, 2019, 6(2): 916-922.
[38] UNIVERSITY B, Y O F, T A J E, et al. Supervised machine learning algorithms: classification and comparison[J]. International Journal of Computer Trends and Technology, 2017, 48(3): 128-138.
[39] YAN H Y, HE Z, GAO C, et al. Investment estimation of prefabricated concrete buildings based on XGBoost machine learning algorithm[J]. Advanced Engineering Informatics, 2022, 54: 101789.
[40] EL MAHDI SAFHI A, DABIRI H, SOLIMAN A, et al. Prediction of self-consolidating concrete properties using XGBoost machine learning algorithm: part 1 workability[J]. Construction and Building Materials, 2023, 408: 133560.
[41] PHILIP S, MARAKKATH N. Compressive strength prediction and feature analysis for GGBS-based geopolymer concrete using optimized XGBoost and SHAP: a comparative study of optimization algorithms and experimental validation[J]. Journal of Building Engineering, 2025, 108: 112879.
[42] YU X R, HU T Y, KHODADADI N, et al. Predictive and experimental assessment of chloride ion permeation in concrete subjected to multifactorial conditions using the XGBoost algorithm[J]. Journal of Building Engineering, 2024, 98: 111041.
[43] HAJIBABAEE P, BEHNOOD A, NGO T, et al. Carbonation depth assessment of recycled aggregate concrete: an application of conformal prediction intervals[J]. Expert Systems with Applications, 2025, 268: 126231.