挠度曲率识别环境温度下的桥梁结构损伤方法研究  

Research on structural damage identification by deflection curvature under ambient temperature

作　　者：黄炎 常军[1] HUANG Yan;CHANG Jun(School of Civil Engineering,SUST,Suzhou 215011,China)

机构地区：[1]苏州科技大学土木工程学院,江苏苏州215011

出　　处：《苏州科技大学学报（工程技术版）》2023年第3期31-38,共8页Journal of Suzhou University of Science and Technology（Engineering and Technology Edition）

摘　　要：针对变化的环境温度引起结构健康监测系统发生误判的问题,提出了一种基于挠度曲率的结构损伤识别方法。利用LSTM神经网络的非线性映射优势,建立温度与挠度的神经网络模型,根据测试温度得到结构无损状态下的挠度及挠度曲率,将测试挠度曲率与无损状态做差值,确定损伤状况,包括损伤时刻、损伤位置和近似损伤程度。通过简支梁与三跨连续梁的数值模拟验证了所提方法的有效性与抗噪性。Aiming at the misjudgment of the structural health monitoring system caused by the changing ambient temperature,a structural damage identification method based on deflection curvature is proposed.Using the nonlinear mapping advantage of LSTM neural network,the neural network model of temperature and deflection is established.The deflection and deflection curvature of the structure under non-destructive state are obtained by the test temperature,and the test deflection curvature is compared with the non-destructive state to determine the damage condition,including the damage time,the damage location and the approximate damage degree.The effectiveness and anti-noise performance of the proposed method are verified by numerical simulation of simply supported beams and three-span continuous beams.

关 键 词：环境温度 损伤识别 损伤程度 LSTM神经网络 挠度 曲率 

分 类 号：TU447[建筑科学—岩土工程]