摘要
为减少测量噪声对网架结构模态识别的不利影响,提出基于变分模态分解(VMD)和奇异值分解(SVD)联合降噪的随机子空间(SSI)模态识别方法.首先,以整体正交系数最小作为优化目标,利用鲸鱼优化算法获取VMD所需的最优K、α参数.基于K、α最优值对响应信号进行VMD后,根据各分量频谱曲线的光滑程度来筛选被激发的有效模态分量.其次,在进一步对有效模态分量进行SVD后,将其叠加以得到降噪后的重构信号.最后,采用SSI法对重构信号进行模态识别.某三自由度体系的仿真分析和某3m×3m平板网架的模型试验均表明,本文方法具有良好的降噪效果,可以筛选有效模态分量和提高模态识别精度.
In order to reduce the harmful influence of measurement noise on the modal identification of space trusses,a method of identifying modal parameters based on signal denoising by both the variational mode decomposition(VMD)and the singular value decomposition(SVD)was proposed.Firstly,with the global orthogonal coefficient set as the objective function,the whale optimization method was utilized to find the best K andαwhich should be preset in the VMD.After the VMD for response signals was done based on the best K andα,the smoothness level of frequency spectrum was evaluated to select the effectively triggered mode functions.Secondly,the effectively triggered mode functions were further processed by the SVD,with the de-noised results superimposed to obtain the reconstructed signals.Finally,the time-domain or frequency-domain methods were used in modal identification by utilizing the reconstructed signals.The numerical simulation of a three-degree-of-freedom system and the experimental study of a 3m×3m flat space truss prove that the proposed method performs well in signal denoising,effectively selects triggered mode functions and improves modal identification accuracy.
作者
程润辉
伍晓顺
苗峰
邹韬
冯隆琨
CHENG Run-hui;WU Xiao-shun;MIAO Feng;ZOU Tao;FENG Long-kun(School of Civil and Surveying&Mapping Engineering(Nanchang),Jiangxi University of Science and Technology,Nanchang 330013,China;Shanghai Construction No.4(Group)Co.,Ltd.,Shanghai 201103,China)
出处
《空间结构》
CSCD
北大核心
2022年第4期16-23,63,共9页
Spatial Structures
基金
国家自然科学基金项目(51868026)
江西省自然科学基金项目(20202BAB204028)
江西理工大学校级研究生创新专项资金项目(XY2021-S167)
关键词
网架结构
模态识别
变分模态分解
奇异值分解
信号处理
space truss
modal identification
variational mode decomposition
singular value decomposition
signal processing
