摘要
纤维增强复合材料(FRP)作为一种新型的增强加固材料,由于其强度高、质量轻、防腐蚀、耐疲劳、与混凝土粘结性能好以及便于施工等诸多优点,在混凝土结构修复加固领域得到了广泛的应用。近年来,随着人工智能(AI)的逐渐兴起,机器学习(ML)作为实现AI的一种途径,在水利、建筑等各行各业也得到了长足的发展。首先简单介绍了ML的基本原理,并通过对ML在混凝土结构工程中应用的系统回顾与总结,指出了传统试验和数值模拟分析中FRP增强混凝土断裂研究存在的一些难点和局限性,阐述了基于ML的人工神经网络(ANN)方法在处理混凝土结构问题中的优越性,认为采用ANN方法能够有效解决FRP增强混凝土断裂研究中难以解决的问题;其次,对ANN方法应用于FRP增强混凝土断裂韧度预测中的新思路进行了详细介绍,给出了ANN方法应用于FRP增强混凝土断裂韧度预测的具体流程,并对其流程中的一些步骤给出了建议;最后,对ML应用于FRP增强混凝土断裂方向的深入研究进行了展望,提出了ML应用于FRP增强混凝土断裂方向深入研究的相关问题。
Fiber Reinforced Polymer(FRP),as a new type of reinforcement material,has been widely used in the field of concrete structure repair and reinforcement due to its high strength,light weight,resistance to corrosion and fatigue,effective bonding with concrete,and ease of construction.As artificial intelligence(AI)emerges,machine learning(ML)has become a popular method for its implementation in the water and construction industries in recent years.First of all,the basic principle of ML is briefly introduced in this paper,and by the systematic review and summary of ML application in concrete structure engineering.Some difficulties and limitations of FRP reinforced concrete fracture research in traditional experiment and numerical simulation analyses are highlighted.The superiority of ML-based artificial neural network(ANN)methods in dealing with concrete structure problems is elaborated.It is considered that ANN can effectively solve the problems that are difficult to solve in the research area of FRP reinforced concrete fractures.Secondly,the new idea of ANN methods applied in predicting the fracture toughness of FRP reinforced concrete is introduced in detail.The specific process of ANN methods is outlined,and some suggestions are given for certain steps in the process.Finally,the further research in the application of ML for FRP reinforced concrete fracture direction is prospected,and the related problems of ML application in further research in the research area are put forward.
作者
范向前
刘决丁
史晨雨
葛菲
FAN Xiangqian;LIU Jueding;SHI Chenyu;GE Fei(Nanjing Hydraulic Research Institute,State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering,Nanjing Hydraulic Research Institute,Nanjing 210024,China;Cooperative Innovation Center for Water Safety&Hydro Science,Nanjing 210024,China;School of Architectural Engineering,Tianjin University,Tianjin 300350,China)
出处
《防灾减灾工程学报》
CSCD
北大核心
2023年第3期626-636,共11页
Journal of Disaster Prevention and Mitigation Engineering
基金
国家自然科学基金项目(52171270,51879168,51679150)
黄河水科学研究联合基金(U2243223)资助。