基于数据驱动的大体积结构裂缝深度反演  

CRACK DEPTH DETECTION OF MASSIVE STRUCTURES BASED ON DATA-DRIVEN ALGORITHM

作　　者：江守燕[1] 杜成斌[1] 孙立国[1] JIANG Shou-yan;DU Cheng-bin;SUN Li-guo(Department of Engineering Mechanics,Hohai University,Nanjing 211100,China)

机构地区：[1]河海大学工程力学系,南京211100

出　　处：《工程力学》2023年第4期215-225,共11页Engineering Mechanics

基　　金：国家自然科学基金项目(52279130);国家重点研发计划项目(2018YFE0122400);中国地球物理学会工程物探检测重点实验室开放研究基金项目(CJ2021GE06)。

摘　　要：裂缝是混凝土结构的主要病害,查明裂缝的深度能够为结构的耐久性和安全性评价提供可靠的信息,但同时也是混凝土结构检测的难点之一。提出了一种基于数据驱动的学习算法,通过考察波经过带缝结构传播的信号预测裂缝深度,采用扩展有限元法(Extended finite element methods,XFEM)和边界吸收层模型模拟了带缝大体积混凝土结构中的波传播过程,将接收点的观测信号和裂缝信息配对。基于人工神经网络的机器学习模型建立了基于XFEM数据集的裂缝深度预测模型。对于含未知裂缝信息的混凝土结构,通过测得的观测点信号,利用建立的机器学习模型实现裂缝深度的实时预测。通过2个数值算例验证了该算法的性能,结果表明所提出的数据驱动算法能够准确预测裂缝深度。Cracks are the main diseases of concrete structures.Finding out the depth of cracks can provide reliable information for the durability and safety evaluation of structures,but it is also one of the difficulties in the detection of concrete structures.A data-driven learning algorithm is proposed to predict the crack depth by investigating the signal of wave propagation through the cracked structure.The wave propagation process in the cracked massive concrete structure is simulated by using the extended finite element methods(XFEM)and the boundary absorbing layer model,and the observation signal of the receiving point is paired with the crack information.Based on the machine learning model of artificial neural network,a fracture depth prediction model based on XFEM datasets is established.For the concrete structure with unknown crack information,through the measured observation point signals,the established machine learning model is utilized to realize the real-time prediction of crack depth.The performance of the algorithm is verified by two numerical examples.The results show that the proposed algorithm can accurately predict the crack depth.

关 键 词：扩展有限元法 边界吸收层 数据驱动 大体积结构 裂缝 

分 类 号：TV331[水利工程—水工结构工程]