基于优化VMD的悬索桥挠度温度效应分离方法  

A Method for Separating Temperature Effect on Deflection of Suspension Bridge Based on Optimized VMD

作　　者：郑彩红 申强[1,3] 邓思勉 何小兵[1,2] ZHENG Cai-hong;SHEN Qiang;DENG Si-mian;HE Xiao-bing(School of Civil Engineering,Chongqing Jiaotong University,Chongqing 400074,China;State Key Laboratory of Mountain Bridges and Tunnels,Chongqing 400074,China;Highway Research Institute,Ministry of Transport,Beijing 100088,China)

机构地区：[1]重庆交通大学土木工工程学院,重庆400074 [2]山区桥梁与隧道国家重点实验室,重庆400074 [3]交通运输部公路科学研究院,北京100088

出　　处：《公路交通科技》2021年第6期88-96,119,共10页Journal of Highway and Transportation Research and Development

摘　　要：为了研究悬索桥实测挠度的温度效应分离,进一步分析悬索桥实测挠度与温度之间的关联性,采用了变分模态分解(variational mode decomposition,VMD)方法对挠度信号进行处理。通过经验模态分解(empirical mode decomposition,EMD)和VMD对仿真信号的对比分析,发现VMD较EMD对噪声有更好的鲁棒性,解决了EMD所存在的模态混叠现象,但仍存在端点效应。提出采用3段交叉信号分解消除VMD分解所存在的端点效应,考虑模态个数、惩罚因子对VMD分解的影响,形成了基于优化VMD的温度效应分离方法。利用优化VMD方法将输入挠度信号分解成一系列具有不同尺度的本征模函数(intrinsic mode function,IMF)。根据各个分解模态的中心频率,结合实际温度变化的周期性判定对应的挠度组分,对由温度引起的挠度分量进行组合,研究了悬索桥不同截面挠度与温度的关联性,相关系数均达到0.9以上。结果表明,该方法能有效地分离信号中不同的频率组分,同时还能提取出挠度信号中的趋势项,通过分析各个模态与实测温度的关联性,证明温度效应主要集中在低频段,悬索桥挠度变化主要受温度影响;在模态个数的选择过程中,发现随着模态个数的增大,低频信息的中心频率逐渐趋于稳定,高频信息被分解得更加精细。验证了该方法可用于桥梁长期健康监测的数据分析及桥梁损伤识别。In order to study the temperature effect separation of measured deflection of suspension bridge,and further analyze the correlation between measured deflection and temperature of suspension bridge,the deflection signal is processed by using VMD method.Through the comparative analysis of empirical mode decomposition(EMD)and VMD on the simulation signal,it is found that VMD is more robust to noise than EMD,which solved the modal mixing phenomenon that exists in EMD,but it still exists end effect.The end effect in VMD decomposition is eliminated by the proposed 3-segment cross-signal decomposition.Considering the influence of number of modes and penalty factor on VMD decomposition,the temperature effect separation method based on optimized VMD is formed.The input deflection signal is decomposed into a series of intrinsic mode functions(IMF)with different scales by using the optimized VMD method.According to the central frequency of each decomposed mode and the periodicity of actual temperature change,the corresponding deflection components are determined,and the deflection components caused by temperature are combined to study the correlation between the deflections of different sections of suspension bridge and temperature,the correlation coefficients are all above 0.9.The result shows that(1)The method can effectively separate different frequency components in the signal and extract the trend term in the deflection signal.Through analysing the correlation between each mode and measured temperature,it is proved that the temperature effect is mainly concentrated in the low frequency band,and the deflection of the suspension bridge is mainly affected by temperature.(2)In the process of selecting the number of modes,it is found that with the increase of number of modes,the center frequency of low-frequency information tends to be stable gradually,while the high-frequency information is decomposed more finely.It is verified that the method can be used for data analysis of long-term health monitoring and damage identifi

关 键 词：桥梁工程 信号分离 变分模态分解 挠度 温度效应 

分 类 号：U441.4[建筑科学—桥梁与隧道工程]