摘要
针对传统吊索疲劳寿命计算方法计算效率低、无法考虑交通量增长的问题,基于最小二乘支持向量机(LSSVM)和车辆荷载监测数据进行悬索桥吊索疲劳寿命预测研究。以某大跨度双塔单跨悬索桥为背景,采用LSSVM建立吊索疲劳损伤与车辆荷载监测数据的相关性模型,建模过程中考虑LSSVM模型输入与输出的最优模式以及训练数据长度;建立1根吊索(以29号吊索为例)与其它吊索的日疲劳损伤之间的相关性模型,预测其它吊索的疲劳损伤;考虑日车流量和等效车总重的增长,进行吊索疲劳寿命预测。结果表明:对于29号吊索的4种LSSVM模型,模型Ⅳ的边界条件较其它3种模型更为合理,测试数据的平均相对误差低于模型Ⅰ~Ⅲ;该方法将日疲劳损伤与车辆荷载监测数据进行直接关联;LSSVM相关性模型的预测能力依赖于训练样本的数量,当训练数据长度为284 d时,模型Ⅳ的预测能力较强,其平均相对误差低于5.5%;同时考虑日车流量和等效车总重增长时,疲劳累积损伤显著增长。
This paper presents a method based on least squares support vector machine(LSSVM)and monitored vehicle loads to predict the fatigue life of hanger cables of suspension bridges,which features higher calculation efficiency than the conventional methods and considers the traffic growth issue.The correlation models of hanger cable fatigue damages and monitored vehicle loads of a long-span,two-tower suspension bridge with hanger cables arranged in one span were established,using LSSVM,in which the optimal input and output modes of LSSVM models as well as training data lengths are considered.The correlation models of the daily fatigue damage of one hanger cable(taking the hanger cable No.29 as the case)and those of other hanger cables were built up,to predict the fatigue damages of other hanger cables.On the basis of the daily traffic volume and total weight of equivalent vehicles,the fatigue life of the hanger cable is predicted.As per the analysis,among the four LSSVM models of hanger cable No.29,the boundary conditions of modelⅣare more reasonable than those of the other three models,and the average relative error of the monitored data is lower than modelsⅠ-Ⅲ.The method presented in this paper allows the daily fatigue damage to be directly correlated with the monitored vehicle loads.The prediction capacity of the LSSVM correlation model hinges on the amount of training samples,for a data length of 284 d,the prediction capacity of modelⅣis further strengthened,with an average relative error lower than 5.5%.Considering both the increase of daily traffic flow and total weight of equivalent vehicles,the cumulative fatigue damage significantly increases.
作者
曾国良
邓扬
ZENG Guoliang;DENG Yang(School of Civil Engineering and Architecture,Hunan University of Arts and Science,Changde 415000,China;School of Civil Engineering,Changsha University of Science&Technology,Changsha 410114,China;School of Civil Engineering and Transportation Engineering,Beijing University of Civil Engineering and Architecture,Beijing 100044,China)
出处
《桥梁建设》
北大核心
2025年第1期41-48,共8页
Bridge Construction
基金
国家自然科学基金项目(51878027)
湖南文理学院科技创新团队项目(2020-26-5)。
关键词
悬索桥
吊索
结构健康监测
车辆荷载
疲劳损伤
疲劳寿命
最小二乘支持向量机
相关性模型
suspension bridge
hanger cable
structural health monitoring
vehicle load
fatigue damage
fatigue life
least squares support vector machine
correlation model
