EEMD模态分解算法在振动数据噪声抑制中的应用  

Application of EEMD Modal Decomposition Algorithm in Noise Suppression of Vibration Data

作　　者：李宏伟[1,2] 曾金艳[1,2] 任瑞国 LI Hongwei;ZENG Jinyan;REN Ruiguo(Shanxi Earthquake Agency,Taiyuan,Shanxi 030021,China;National Continental Rift Valley Dynamics Observatory of Taiyuan,Taiyuan,Shanxi 030025,China)

机构地区：[1]山西省地震局,山西太原030021 [2]太原大陆裂谷动力学国家野外科学观测研究站,山西太原030025

出　　处：《山西地震》2024年第4期40-45,共6页Earthquake Research in Shanxi

基　　金：山西省自然科学研究面上项目(202303021221251);山西省地震局科研项目(SBK-2319)。

摘　　要：桥梁结构响应容易受到环境噪声干扰,导致振动响应信号中的各模态分量无法准确提取。针对现有降噪方法的局限性,引入EEMD经验模态分解算法,对桥梁振动信号进行降噪处理,基于数学模拟实验分析EEMD降噪效果。结果表明,该方法能较好地过滤叠加在已知曲线上的高斯白噪声,且大幅度提高信噪比。在实测的振动观测曲线上,EEMD降噪较好地消除了混杂在实测振动数据曲线中的随机误差和异常波动成分;降噪后的振动曲线更加平稳,FFT频谱曲线能直观地反映出更清晰、连续的数据。The response of bridge structure is easily disturbed by environmental noise,which leads to the failure to extract the modal components of vibration response signal accurately.In response to the limitations of existing noise reduction methods,this paper introduces the EEMD empirical mode decomposition algorithm to denoise bridge vibration signals.Mathematical simulation experiments were conducted to analyze the effectiveness of EEMD in noise reduction.The results indicate that this method effectively filters out Gaussian white noise superimposed on known curves and significantly enhances the signal-to-noise ratio.In actual measured vibration curves,EEMD denoising successfully removes random errors and abnormal fluctuations present in the data,resulting in a smoother vibration curve.Additionally,the FFT spectrum of the denoised data provides a clearer and more continuous representation of the signal.

关 键 词：桥梁结构 EEMD降噪方法 振动数据 

分 类 号：TU352.1[建筑科学—结构工程]