基于ICA的工作模态参数识别

Operational Modal Parameter Identification Based on ICA

本文分别阐述了独立分量分析和基于ICA的工作模态分析原理,发现了ICA分离模型与结构振动模态分析模型的一致性。应用ICA算法和比利时LMS公司的OMA分析软件分别对齿轮箱正常和断齿工况进行模态参数识别,对比发现,ICA算法与目前最常用的Op.Poly MAX算法相比抗噪性强,识别简便精准,为工作模态参数识别提供新的识别依据。

The paper expounds the principle of independent component analysis and the operational modal analysis based on the principle of ICA. The analysis demonstrates the consistency between ICA separation model and structural vibration modal analysis model. The ICA algorithm and the software OMA developed by Belgian LMS are applied to identify the condition of gear box and broken tooth through modal parameters respectively. This paper finds that the algorithm of ICA, comparing with Op.PolyMAX which is the most commonly used, has strong anti-noise performance. Besides, it is easy to operate and the identification is accurate, this kind of algorithm provides a new basis of work modal parameters identification.