nav emailalert searchbtn searchbox tablepage yinyongbenwen piczone journalimg journalInfo journalinfonormal searchdiv searchzone qikanlogo popupnotification paper paperNew
2025, 02, v.46 112-119
基于BO-KELM的大跨变截面连续箱梁桥损伤识别研究
基金项目(Foundation): 河南省科技攻关项目(202102310251); 中建七局科技研发课题(CSCEC7b-2021-Z-4); 华北水利水电大学研究生创新能力提升工程(NCWUYC-2023046,NCWUYC-202315058)
邮箱(Email): 84175817@qq.com;
DOI: 10.19760/j.ncwu.zk.2025029
发布时间: 2024-07-03
出版时间: 2024-07-03
网络发布时间: 2024-07-03
移动端阅读
摘要:

大跨度变截面连续箱梁桥是我国公路和市政桥梁中经常采用的典型结构型式,在运营阶段,精准识别桥梁损伤是准确评价其结构健康状态的首要条件。针对桥梁缺陷产生的不确定性,提出了一种基于BO-KELM识别大跨度变截面连续箱梁桥损伤的方法。试验和工程应用结果表明:放大曲率模态作为损伤识别指标,可有效识别大型结构单点和多点的轻微损伤位置;基于KELM构建的预测模型并结合BO算法,可快速识别梁桥不同程度的损伤,预测精度可达98.75%,较传统神经网络提高10%。提出的放大曲率模态和BO-KELM损伤识别方法具有对微小损伤识别能力强、损伤程度判别速度快、精度高等特点。研究结果可为大跨度变截面连续箱梁桥损伤位置和损伤程度的快速精准识别提供理论和方法指导。

Abstract:

Long-span variable cross-section continuous box-girder bridges represent a typical structural form widely used in Chinese highway and municipal bridges. During operational phases, accurate damage identification serves as the prerequisite for precise structural health assessment. To address uncertainties in bridge defects, this study proposes a BO-KELM(Bayesian Optimization-Kernel Extreme Learning Machine) based method for damage identification in such bridges. Experimental and engineering application results demonstrate that:(1) The amplified curvature mode, as a damage identification index, effectively detects locations of minor single-point and multi-point damage in large-scale structures;(2) The prediction model constructed using KELM combined with BO algorithm enables rapid identification of varying damage severities, achieving 98.75% prediction accuracy—a 10% improvement over traditional neural networks. The proposed methodology exhibits distinctive advantages including enhanced sensitivity to minor damage, rapid damage severity assessment, and high precision. These research findings provide valuable references for fast and accurate identification of damage locations and severities in long-span variable cross-section continuous box-girder bridges.

参考文献

[1] TU Y M,LU S L,WANG C.Damage identification of steel truss bridges based on deep belief network[J].Journal of Southeast University,2023,38(4):392-400.

[2] 刘博楠.基于贝叶斯网络的大跨径悬浇拱桥施工风险分析研究[D].重庆:重庆交通大学,2021.

[3] 熊峰.大跨径连续梁桥下挠及其对策的研究[D].长沙:湖南大学,2018.

[4] 邹宇,徐栋,宋冰泉,等.预制节段桥梁钢榫键接缝受力特性[J].同济大学学报(自然科学版),2021,49(7):986-994.

[5] LEE S G,YUN G J,SHANG S.Reference-free damage detection for truss bridge structures by continuous relative wavelet entropy method[J].Structural Health Monitoring,2014,13(3):307-320.

[6] OBRIEN E J,FITZGERALD P C,MALEKJAFARIAN A,et al.Bridge damage detection using vehicle axle-force information[J].Engineering Structures,2017,153:71-80.

[7] 周云,蒋运忠,易伟建,等.基于模态柔度理论的结构损伤诊断试验研究[J].湖南大学学报(自然科学版),2015,42(5):36-45.

[8] 杨则英,张亚磊,曲建波,等.基于BP神经网络的中等跨径桥梁损伤识别探讨[J].公路与汽运,2020(3):136-140,155.

[9] 王静,张伟,王骑.基于曲率模态法的简支板桥损伤识别研究[J].工业建筑,2006(增刊1):225-227.

[10] 陈笑,胡宏祥,戚王月,等.基于集对分析和GA-BP神经网络的地下水埋深预测研究[J].华北水利水电大学学报(自然科学版),2019,40(4):57-64.

[11] 刘汉东,张世英.基于非线性分析法的碎裂结构岩体力学参数预测[J].华北水利水电大学学报(自然科学版),2020,41(6):78-84.

[12] 项长生,原子,周宇等.基于曲率模态信息熵和BP神经网络的简支梁损伤识别方法[J].沈阳工业大学学报,2022,44(5):575-583.

[13] 徐宏文,李宁,谢永健.基于模态曲率曲线拟合的板结构损伤识别研究[J].工程抗震与加固改造,2018,40(4):16-20,26.

[14] 吴桐,唐亮,周志祥.基于曲率模态面积差方比的桥梁结构损伤识别[J].公路交通科技,2021,38(11):59-67.

[15] PANDEY A K,BISWAS M.Experimental verification of flexibility difference method for locating damage in structures[J].Journal of Sound and Vibration,1995,184(2):311-328.

[16] 刘洪公,王学军,李冰莹,等.基于卷积神经网络的桥梁裂缝检测与识别[J].河北科技大学学报,2016,37(5):485-490.

[17] 刘军香,王立新,姜慧,等.基于加速度二次协方差矩阵和神经网络的结构损伤识别[J].地震工程与工程振动,2019,39(3):214-221.

[18] 许如锋,陈建国,赵作周,等.基于应变和PSO-BP神经网络的连续梁桥损伤识别方法[J].安全与环境学报,2017,17(6):2093-2098.

[19] 陈震,王亚茹,陈璐,等.基于学习因子优化的粒子群算法识别结构损伤[J].华北水利水电大学学报(自然科学版),2022,43(4):43-47,75.

[20] 诸德熙,邓华,王玮.保持承载刚度的张拉整体结构预应力降低方法[J].建筑结构学报,2022,43(4):95-102.

[21] 时宏森,杨涛,唐超,等.基于ANSYS Workbench的扭转弹簧疲劳寿命分析[J].新技术新工艺,2021,401(5):24-29.

[22] 左帆.基于混凝土界面过渡区非均匀水化特点的杨氏模量跨尺度计算模型研究[D].西安:西安建筑科技大学,2022.

[23] 任宜春,陈义龙.呼吸裂缝梁基于奇异谱熵的损伤识别方法研究[J].应用力学学报,2019,36(1):230-234.

[24] 朱岩,蒋青桔.曲率模态参数法在张弦结构定量损伤识别中的研究和应用[J].工业建筑,2022,52(10):71-77.

[25] 彭华春,张康康,时松,等.节段预制拼装桥梁研究综述[J].铁道标准设计,2022,66(10):75-83.

[26] 魏锦辉,潘春风,张青阔.简支梁多位置损伤识别研究[J].华北水利水电学院学报,2009,30(4):37-39.

[27] LU W L,LI F L,ZHU L,et al.Experimental study on the shear behavior of precast concrete segmental simply supported beam with internal tendons and glue joints[J].Case Studies in Construction Materials,2022,17:e01545.

[28] FAN Y Y,WANG H C,ZHAO X Y,et al.Short-term load forecasting of distributed energy system based on kernel principal component analysis and KELM optimized by fireworks algorithm[J].Applied Sciences,2021,11(24):12014.

[29] LIM J,CHOI M,LEE S.A Bayesian optimization algorithm for the optimization of mooring system design using time-domain analysis[J].Journal of Marine Science and Engineering,2023,11(3):507.

[30] 卢山.基于曲率模态的钢筋混凝土梁损伤识别的理论分析与试验研究[D].武汉:武汉理工大学,2019.

基本信息:

DOI:10.19760/j.ncwu.zk.2025029

中图分类号:U446

引用信息:

[1]吴庆亮,赵洋,崔书宇,等.基于BO-KELM的大跨变截面连续箱梁桥损伤识别研究[J].华北水利水电大学学报(自然科学版),2025,46(02):112-119.DOI:10.19760/j.ncwu.zk.2025029.

基金信息:

河南省科技攻关项目(202102310251); 中建七局科技研发课题(CSCEC7b-2021-Z-4); 华北水利水电大学研究生创新能力提升工程(NCWUYC-2023046,NCWUYC-202315058)

发布时间:

2024-07-03

出版时间:

2024-07-03

网络发布时间:

2024-07-03

检 索 高级检索