摘要
为解决传统风险识别方法在复杂的桥梁风险情景下识别能力不足的问题,以某三跨连续梁桥为研究对象,构建该桥多层次风险情景库。基于改进的深度前馈神经网络(DFNN),对桥梁的不同风险类型、风险位置及风险程度进行特征学习,并用训练完毕的网络对典型风险情景进行预测分析,实现桥梁多层次风险识别研究。结果表明,加入L2正则化可以显著提升DFNN的泛化能力,使分类、定位及定量识别的训练准确度分别达到91.6%、74.9%和93.7%;网络在有限元模拟典型风险识别任务中表现出较高的预测准确率,分别为80%、100%和97.41%。总体而言,提出的L2-DFNN可以较好地实现桥梁多层次风险识别工作,可为桥梁健康监测(BHM)提供重要支撑。
A three-span continuous girder bridge was studied to establish a multilevel risk scenario library and address the insufficient identification capability of traditional risk assessment methods under complex and unknown bridge risk scenarios.Feature learning was performed on different risk types,locations,and severity levels of bridges using an improved deep feedforward neural network(DFNN).The trained network was used to predict and analyze typical risk scenarios,achieving multilevel risk identification for bridges.Results demonstrate that L2 regularization significantly enhances the generalization capability of DFNN,yielding training accuracies of 91.6%for classification,74.9%for localization,and 93.7%for quantitative identification.In the finite element simulations of typical risk identification tasks,the prediction accuracies reach 80%,100%,and 97.41%,respectively.The proposed L2-DFNN effectively enables multilevel bridge risk identification,providing critical support for bridge health monitoring.
作者
张在羽
李舒
甘露一
刘洪涛
ZHANG Zaiyu;LI Shu;GAN Luyi;LIU Hongtao(Hefei Institutefor Public Safety Research,Tsinghua University,Hefei 230601,Anhui,China;Anhui University of Science and Technology,Huainan 232001,Anhui,China;Anhui Provincial Key Laboratory of Intelligent Detection and Diagnosis for Traffic Infrastructure,Hefei 230051,Anhui,China)
出处
《地震工程学报》
2025年第6期1357-1368,共12页
China Earthquake Engineering Journal
基金
国家重点研发计划项目(2024YFC3016800)
安徽省科技攻坚计划项目(202423l10050030)
安徽理工大学硕士研究生创新基金项目(2023CX2026)。
关键词
多层次风险识别
深度前馈神经网络
L2正则化
有限元模拟
桥梁健康监测
multilevel risk identification
deep feedforward neural network(DFNN)
L2 regularization
finite element simulation
bridge health monitoring
作者简介
第一作者:张在羽(1999-),男,硕士研究生,主要从事桥梁抗震与安全监测、桥梁风险预警方面研究。E-mail:zhangyouduowen@163.com;通信作者:李舒(1984-),男,副研究员,硕士生导师,主要从事桥梁抗震与安全监测、城市生命线防灾减灾工程研究。E-mail:lishu@tsinghua-hf.edu.cn。
