桥梁结构多测组振型融合的两阶段快速贝叶斯方法研究  

A Double-stage Fast Bayesian Approach for Mode Shape Assembly of Bridges with Multiple Test Setups

作　　者：刘玲玲 颜王吉 李丹[1] 任伟新 秦超 LIU Lingling;YAN Wangji;LI Dan;REN Weixin;QIN Chao(College of Civil andWater Conservancy Engineering,Hefei University of Technology,Hefei 230009,China;State Key Laboratory of Internet of Things for Smart City,University of Macao,Macao 999078,China;Department of Civil and Environmental Engineering,University of Macao,Macao 999078,China;College of Civil and Transportation Engineering,Shenzhen University,Shenzhen 518060,Guangdong,China)

机构地区：[1]合肥工业大学土木与水利工程学院,合肥230009 [2]澳门大学智慧城市物联网国家重点实验室,澳门999078 [3]澳门大学土木与环境工程系,澳门999078 [4]深圳大学土木与交通工程学院,广东深圳518060

出　　处：《噪声与振动控制》2021年第4期182-189,197,共9页Noise and Vibration Control

基　　金：国家自然科学基金资助项目(51778203,51778204,51708164)。

摘　　要：桥梁振动实验常需进行分组多次测试,如何从各测组数据中准确识别频率、阻尼比、局部振型,进而将局部振型组装成整体振型至关重要。结构模态分析过程不可避免地受到测试噪声和模型误差的影响,引入贝叶斯方法进行模态参数识别和振型融合的不确定性量化分析具有重要意义。提出桥梁结构多测组振型融合的两阶段快速贝叶斯方法。第一阶段采用快速贝叶斯快速傅里叶变换(FFT)方法求解各测组频率、阻尼比和局部振型的最优值及协方差;第二阶段利用局部振型协方差信息自动分配各测组的权重,基于贝叶斯原理形成负对数似然函数,通过解析耦合迭代优化算法快速得到整体振型的最优值,并推导出负对数似然函数关于整体振型的Hessian矩阵的解析解,用于量化整体振型的不确定性。最后,通过一座斜拉桥的环境激励振动测试数据验证该方法的有效性,并研究各测组数据质量及频带宽度等因素对整体振型不确定性的影响。分析表明,该方法计算效率高,并能合理量化不确定性。Dynamic testing for bridges usually needs to be divided into several setups due to the limitation of sensors.Therefore,precise identification of frequency,damping ratio and local mode shapes from different setups is very important for assembling local mode shapes to global mode shapes.In the process of structural modal analysis,the disturbances of testing noise and model errors are inevitable.Thus,uncertainty quantification has attracted widespread attention.In this study,a double-stage fast Bayesian approach for mode shape assembly were proposed to process the dynamic testing data of a bridge.First of all,the optimal values and the mean square differences of the natural frequencies,damping ratios and local mode shapes were identified for different setups using fast Bayesian FFT approach.Then,local mode shapes were glued together to form the global mode shapes,in which the optimal values and their corresponding uncertainties were computed by analytical algorithms.The effects of data quality and bandwidth on the results were also investigated.Results show that the fast Bayesian approach can achieve the global mode shapes effectively and quantify their uncertainties properly.

关 键 词：振动与波 桥梁工程 模态参数识别 振型融合 不确定性 贝叶斯方法 

分 类 号：U441.3[建筑科学—桥梁与隧道工程]