引用本文:芦竹茂,白洋,黄纯德,关少平,孟晓凯.变分模态分解组合广义形态滤波器的MEMS陀螺仪去噪方法[J].控制理论与应用,2023,40(3):509~515.[点击复制]
LU Zhu-mao,BAI Yang,HUANG Chun-de,GUAN Shao-ping,MENG Xiao-kai.De-noising method of MEMS gyroscope based on variational mode decomposition combined generalized morphological filter[J].Control Theory and Technology,2023,40(3):509~515.[点击复制]
变分模态分解组合广义形态滤波器的MEMS陀螺仪去噪方法
De-noising method of MEMS gyroscope based on variational mode decomposition combined generalized morphological filter
摘要点击 969  全文点击 278  投稿时间：2021-03-31  修订日期：2022-05-25
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2021.10272
  2023,40(3):509-515
中文关键词  变分模态分解  组合广义形态滤波  结构元素  MEMS陀螺仪  微机电系统  信号去噪
英文关键词  VMD  CGMF  SE  MEMS gyroscope  microelectromechanical systems  signal denoising
基金项目  国网山西省电力公司科技项目(52053018000T)资助.
作者单位邮编
芦竹茂* 国网山西省电力公司电力科学研究院 030001
白洋 国网山西省电力公司电力科学研究院 
黄纯德 国网山西省电力公司电力科学研究院 
关少平 国网山西省电力公司 
孟晓凯 国网山西省电力公司电力科学研究院 
中文摘要
      为了更加有效地消除MEMS陀螺仪输出信号存在大量不同类型噪声的同时保留有效信号特征, 本文提出 了一种变分模态分解(VMD)的多尺度自适应组合广义形态滤波器(CGMF)去噪方法. 该方法首先采用VMD将MEMS 陀螺仪原始输出信号分解为多个不同尺度的具有特殊稀疏性的一高低频离散带限子信号内模函数(BLIMFs), 然后通过选择CGMF中合适的结构元素(SEs)长度和几何结构对上述不同尺度BLIMFs进行自适应去噪处理, 最后重 建去噪后的BLIMFs获得去噪信号. 通过实验验证并与现有的信号去噪方法相比, 本方法的主要优点在于: 1) 解决 了CGMF中SEs的长度和几何结构等关键参数的自适应选择问题; 2) 针对不同类型噪声均进行了有效的分离和去噪 处理.
英文摘要
      In order to effectively eliminate a large number of different types of noise in the output signal of the MEMS gyroscope while preserving the effective signal characteristics, a multi-scale adaptive combined generalized morphological filter (CGMF) denoising method based on the variational mode decomposition (VMD) is proposed in this paper. Firstly, the original output signal of the MEMS gyroscope is decomposed into a number of high and low frequency discrete band limited intrinsic mode functions (BLIMFs) of different scales with special sparsity by VMD. Then, the adaptive denoising is performed on the BLIMFs of different scales by selecting appropriate structural elements (SEs) length and geometric structure in CGMF. Finally, the denoised BLIMFs is reconstructed to obtain the denoised signal. Compared with the existing signal denoising methods, the main advantages of this method are as follows: 1) it solves the adaptive selection of key parameters such as the SEs length and geometric structure in CGMF; 2) effective separation and denoising are carried out for different types of noise.