基于随机子空间法的模态参数自动提取

Autonomous Modal Parameter Extraction Based on Stochastic Subspace Identification

为了自动且有效地剔除虚假模态、避免人为参与模态模型定阶及模态拾取,进而实现模态参数的自动准确识别,提出了一种新型的随机子空间法运行模态参数自动识别方法。该方法首先构造2个不同维度的汉克尔矩阵对应的随机子空间法运行模态识别模型并求解其极点,在此基础上针对2个模型的同阶极点进行匹配进而获得超清的稳态图,再对该稳态图进行谱系聚类分析,最终自动准确识别出模态参数。5自由度质量-弹簧-阻尼仿真系统及矩形平板的运行模态试验识别结果均验证了方法的有效性。

To remove spurious modes automatically and effectively, to avoid artificial participation in modal model order determination and modal pickup, so as to extract the modal parameter automatically and exactly, a novel autonomous modal parameter extraction method based on stochastic subspace identification method is proposed. Firstly, two stochastic subspace models with different dimension Hankel matrices for operating modal parameter identification are constructed and their poles are computed. On this basis, a crystal stabilization diagram is obtained by matching the same order poles of the two models. Finally, hierarchical clustering analysis is performed to the stabilization diagram and then exact modal parameters are automatically extracted. The effectiveness of the proposed method is verified by the simulation of operational modal analysis for a 5 DOF mass-spring-damping system and the experiment for a rectangle plate.