基于随机子空间法的结构模态参数自动识别方法

Automatic identification method for structural modal parameters based on stochastic subspace identification

为解决随机子空间法在稳定点自动分析方面存在的抗噪能力不足问题,提出了一种新的自动化识别方法.首先,采用协方差驱动的随机子空间法,按照新的定义方式输出稳定点.其次,采用改进的OPTICS算法对稳定点进行清洗、聚类.然后,采用基于频率中值的自适应合并方法,对未完全合并的簇进行有效聚合,并以簇中值作为模态参数的代表值,实现完全无人工干预的自动化模态识别.最后,以Lysefjord悬索桥模型为例进行模态识别,验证该方法的可行性.结果表明:所提方法在实现自动化的同时,具有较高的精度,频率值最大误差仅为1.926%,且在各程度噪声干扰下都能以较高的成功率自动、准确地识别模态参数,相比于对照方法,其鲁棒性优势明显.

To solve the problem of insufficient anti-noise ability in the automatic analysis of stable poles by stochastic subspace identification(SSI) method, a new automatic identification method for modal parameters was proposed. Firstly, stable poles were output by covariance driven stochastic subspace identification(COV-SSI) combined with a new definition of stable poles. Secondly, the modified ordering points to identify the clustering structure(OPTICS) algorithm was used to clean and cluster stable poles. Thirdly, an adaptive merging method based on the median frequency was proposed to aggregate the incompletely merged clusters, and the cluster median was used as the representative value of the modal parameters to realize automatic modal identification without manual intervention. Finally, the feasibility was validated by taking the Lysefjord suspension bridge model as an example. The results show that the proposed method can achieve automation with high accuracy, and the maximum error of the frequency value is only 1.926%. It can automatically and accurately identify the modal parameters at various levels of noise interference, and its robustness advantage is obvious compared with the control methods.