基于AR-EWT和SSI的桥梁结构模态参数识别

Modal identification for bridge structures using AR-EWT-based SSI technique

为解决随机子空间识别(SSI)算法存在的系统定阶困难问题,避免各模态之间的相互影响,提出一种结合自回归功率谱经验小波变换(AR-EWT)和SSI的桥梁结构模态参数识别方法。首先,采用AR-EWT将结构不同测点的振动响应信号分解成一系列仅含单一模态信息的单分量信号,并利用奇异值分解(SVD)去噪单分量信号。然后,组装所有测点中包含相同模态信息的单分量信号。最后,利用SSI分别处理各组单分量信号,识别结构的自振频率、阻尼比和振型。ASCE Benchmark模型、简支梁模型和西宁北川河桥的模态参数识别结果表明:所提出的AR-EWT结合SSI的模态参数识别方法可避免直接利用SSI识别模态参数时存在的系统定阶问题,提高模态参数识别效率;ASCE Benchmark模型、简支梁模型和西宁北川河桥的模态参数识别结果与直接利用SSI的识别结果、有限元分析结果基本一致。

To resolve the difficulty of the model order determination in stochastic subspace identification(SSI)and to avoid the interaction among vibration modes,a modal parameter identification method for bridge structures by using the autoregressive power spectrum-based empirical wavelet transform(AR-EWT)and SSI was proposed.Firstly,the measured structural vibration responses at different measurement points were decomposed into a series of mono-component signals using AR-EWT,and then the singular value decomposition(SVD)was adopted to denoise the mono-component signals.Then,the mono-component signals containing the same modal information were assembled into a group.Finally,each group of mono-component signal was processed using SSI,and the natural frequencies,damping ratios and mode shapes of the bridge structures were identified.The results show that the proposed AR-EWT-based SSI technique can avoid the model order determination problem in SSI,and improve the efficiency of modal parameter identification.The identified modal parameter results of the ASCE Benchmark model,the simply supported beam model,and the Xining Beichuan River Bridge using the proposed AR-EWTbased SSI technique are consistent with the identified ones using SSI technique and the finite element analysis results.