Structural identification based on substructure decoupling technique considering uncertain dynamic model

doi:10.1016/j.ymssp.2024.111957

Structural identification based on substructure decoupling technique considering uncertain dynamic model

2025-02-01

0

PRE

AI

DOI：10.1016/j.ymssp.2024.111957

原文链接

原文求助

摘要

En 中文

Structural identification based on substructure decoupling technique (SDT) is an efficient and widely applied method. The input for SDT consists of the frequency response functions (FRFs) obtained from testing and dynamic model solutions of the assembly and substructure. Consequently, one of the significant factors impacting identification accuracy is the error in the dynamic model. A structural identification method considering uncertain dynamic model is proposed in this work, taking into account the inherent uncertainty of these errors. The structural receptance is treated as a random matrix, and a distribution model is formulated to quantify its stochastic characteristics. The distribution model comprehensively accounts for the randomness of structural modal parameters. Through this method, random receptances are sampled directly in the matrix, instead of in the numerous modal parameters in traditional method, which effectively reduces the dimensionality of randomness. The identified results are obtained through a Monte Carlo simulation (MCS) process, whose results constitute an interval. The application of the proposed method is not limited by specific structural forms or dynamic modeling approaches. Compared with traditional identification method that does not consider uncertainty, only three additional coefficients need to be introduced, showing its nice practicality. A simulation and an experimental study are conducted for validation. The results indicate that random model errors in the dynamic model have a substantial impact on identification accuracy, while the proposed method can significantly improve the identification performance compared to traditional method. Additionally, the influencing factors affecting the method's effectiveness are studied, and suggestions to achieve better identifying are provided regarding the selection of measured points.

Keyword:

Structural uncertainty

Structural identification

Model errors

Substructure decoupling

Monte Carlo

Structural identification based on substructure decoupling technique considering uncertain dynamic model

摘要

期刊

机构

引用论文