自整定多元变分模态分解

郎恂; 王佳艺; 陈启明; 何冰冰; 毛汝凯; 谢磊

doi:10.11999/JEIT230763

自整定多元变分模态分解

doi: 10.11999/JEIT230763

1.
云南大学信息学院昆明 650504
2.
浙江大学工业控制技术国家重点实验室杭州 310027
3.
云南云天化股份有限公司装备技术中心昆明 650000

基金项目: 国家自然科学基金(62003298, 62201495)，云南省基础研究计划(202301AT070277)，云南省重大科技专项(202202AD080005, 202202AH080009)

详细信息

作者简介:
郎恂：男，副教授，研究方向为智能信号处理、工业过程智能监测与控制、医疗信号处理与应用等

王佳艺：女，硕士生，研究方向为自适应多元信号处理与时频分析

陈启明：男，高级工程师，研究方向为自适应信号处理与时频分析等

何冰冰：女，讲师，研究方向为医疗信号处理与应用

毛汝凯：男，研究方向为工业过程智能监测与控制

谢磊：男，教授，博士生导师，研究方向为医疗信号处理与深度学习、工业过程智能优化与控制等

通讯作者:
何冰冰　hebingbing@ynu.edu.cn

中图分类号: TN911.7
计量
- 文章访问数: 473
- HTML全文浏览量: 207
- PDF下载量: 48
- 被引次数: 0
出版历程
- 收稿日期: 2023-07-25
- 修回日期: 2024-04-27
- 网络出版日期: 2024-05-15
- 刊出日期: 2024-07-29

Self-tuning Multivariate Variational Mode Decomposition

1.
School of Information Science and Engineering, Yunnan University, Kunming 650504, China
2.
State Key Laboratory of Industrial Control Technology, Zhejiang University, Hangzhou 310027, China
3.
Yunnan Yuntianhua Limited by Share Ltd., Kunming 650000, China

Funds: The National Natural Science Foundation of China (62003298, 62201495), Yunnan Fundamental Research Projects (202301AT070277), The Major Project of Science and Technology of Yunnan Province (202202AD080005, 202202AH080009)

摘要

摘要: 多元变分模态分解(MVMD)作为变分模态分解(VMD)的多元扩展，在继承VMD优点的同时，也存在其分解性能很大程度上依赖于两个预置参数——模态数量K 和惩罚系数 $\alpha$ 的问题。为此，该文提出一种自整定MVMD(SMVMD)算法。SMVMD采取了匹配追踪法的思想，通过频域的能量占比和模态正交性分别自适应地更新K和 $\alpha$ 。对仿真信号与真实案例的分析结果表明，所提SMVMD方法不仅有效解决了原MVMD的参数整定问题，而且表现出以下优势，(1) 与MVMD相比，SMVMD抗模态混叠的能力更强，且对噪声和 $\alpha$ 值的变化都具有更好的鲁棒性。(2) 与多元经验模态分解、快速多元经验模态分解和多元变分模态分解这些经典算法相比，SMVMD算法的分解误差最小，分解效果最好。
- 多元信号处理 /
- MVMD /
- 自整定 /
- 匹配追踪法 /
- 鲁棒性
Abstract: The Multivariate Variational Mode Decomposition (MVMD), being an extension of the Variational Mode Decomposition (VMD), inherits the merits of VMD. However, it encounters an issue wherein its decomposition performance relies heavily on two predefined parameters, the number of modes (K) and the penalty factor ( $\alpha$ ). To address this issue, a Self-tuning MVMD (SMVMD) algorithm is proposed. SMVMD employs the notion of matching pursuit to adaptively update K and $\alpha$ based on energy occupation and mode orthogonality in the frequency domain, respectively. The experimented results of both simulated signals and real cases demonstrate that the proposed SMVMD not only effectively addresses the parameter rectification problem of the original MVMD, but also exhibits the following advantages: (1) SMVMD displays superior resilience to mode-mixing compared to MVMD, along with enhanced robustness to both noise and variations in $\alpha$ -value. (2) In comparison to the classical algorithms of multivariate empirical mode decomposition, fast multivariate empirical mode decomposition, and multivariate variational mode decomposition, SMVMD showcases the lowest decomposition error and the best decomposition effect.
- Multivariate signal processing /
- Multivariate Variational Mode Decomposition (MVMD) /
- Self-tuning /
- Matching pursuit method /
- Robustness

HTML全文

图 1 滤波器组特性

下载: 全尺寸图片幻灯片

图 2 X₁(t)分解的噪声鲁棒性和 $\alpha$ 值鲁棒性

下载: 全尺寸图片幻灯片

图 3 SMVMD分解EEG信号的结果

下载: 全尺寸图片幻灯片

图 4 厂级振荡数据及SMVMD的分解结果

下载: 全尺寸图片幻灯片

图 5 基于SMVMD的厂级振荡检测结果可视化

下载: 全尺寸图片幻灯片

1 SMVMD算法

初始化： $\{ u_{k,c}^1(\omega )\} ,\{ \omega _k^1\} ,n \leftarrow 0$
令 $k = 1,{{\mathbf{r}}_1}(\omega ) = {\mathbf{x}}(\omega )$
while $\left\\| {{{\mathbf{r}}_{{k}}}(\omega )} \right\\|_2^2/\left\\| {{\mathbf{x}}(\omega )} \right\\|_2^2 > \delta$ do
令n=1，初始化中心频率 $\{ \omega _k^1\}$
while $\displaystyle\sum\limits_c {\dfrac{{\|\|u_{k,c}^n(\omega ) - u_{k,c}^{n - 1}(\omega )\|\|_2^2}}{{\|\|u_{k,c}^{n - 1}(\omega )\|\|_2^2}}} \gt \varepsilon$ do
$n \leftarrow n + 1$
for $c = 1:C$ do
$u_{k,c}^{n + 1}(\omega ) \leftarrow \dfrac{{{r_{k,c}}(\omega )}}{{1 + 2\alpha _c^n{{(\omega - \omega _k^n)}^2}}}$
end for
$\omega _k^{n + 1} \leftarrow \dfrac{{\displaystyle\sum\limits_c {\int\limits_0^\infty {\omega \|u_{^{_{k,c}}}^{n + 1}(\omega ){\|^2}{\text{d}}\omega } } }}{{\displaystyle\sum\limits_c {\int\limits_0^\infty {\|u_{k,c}^{n + 1}(\omega ){\|^2}{\text{d}}\omega } } }}$
for $c = 1:C$ do
$\alpha _c^{n + 1} = \alpha _c^n\dfrac{{u_{k,c}^{n + 1}{{(\omega )}^}u_{k,c}^{n + 1}(\omega )}}{{u_{k,c}^{n + 1}{{(\omega )}^}{r_{k,c}}(\omega )}}$
end for
end while
令 ${u_{k,c}}(\omega ) = u_{k,c}^{n + 1}(\omega ),{\omega _k} = \omega _k^{n + 1}$ ，同时更新
${{\mathbf{r}}_{k + 1}}(\omega ) = {\mathbf{x}}(\omega ) - {{\mathbf{u}}_k}(\omega )$
end while

下载: 导出CSV

表 1 4种算法的SAD值

算法	MVMD	MEMD	FMEMD	SMVMD
SAD	858.30	854.64	993.42	226.11

下载: 导出CSV

表 2 归一化相关系数与稀疏指数

	模态	x₁	x₂	x₃	x₄	x₅	x₆	x₇	x₈	x₉
${\theta _{k,c}}$	u₁	0.446	0.497	0.398	0.441	0.371	0.497	1	1	1
${\theta _{k,c}}$	u₂	1	1	1	1	1	1	0.491	0.419	0.493
SI	u₁	–	–	–	–	–	–	0.8936	0.8853	0.8769
SI	u₂	0.9117	0.8818	0.8814	0.8866	0.8361	0.8306	-	-	-

下载: 导出CSV

参考文献(28)

[1]	CHEN Qiming, LANG Xun, XIE Lei, et al. Multivariate intrinsic chirp mode decomposition[J]. Signal Processing, 2021, 183: 108009. doi: 10.1016/j.sigpro.2021.108009.
[2]	ZAHRA A, KANWAL N, REHMAN N U, et al. Seizure detection from EEG signals using multivariate empirical mode decomposition[J]. Computers in Biology and Medicine, 2017, 88: 132–141. doi: 10.1016/j.compbiomed.2017.07.010.
[3]	HAN G, LIN B, and XU Z. Electrocardiogram signal denoising based on empirical mode decomposition technique: An overview[J]. Journal of Instrumentation, 2017, 12(3): P03010. doi: 10.1088/1748-0221/12/03/P03010.
[4]	王宇红, 高志兴. 改进多维本质时间尺度分解的厂级振荡检测[J]. 控制工程, 2022, 29(10): 1835–1840. doi: 10.14107/j.cnki.kzgc.CAC2020-1557. WANG Yuhong and GAO Zhixing. Plant-wide oscillation detection based on improved multivariate intrinsic time-scale decomposition[J]. Control Engineering of China, 2022, 29(10): 1835–1840. doi: 10.14107/j.cnki.kzgc.CAC2020-1557.
[5]	REHMAN N U and AFTAB H. Multivariate variational mode decomposition[J]. IEEE Transactions on Signal Processing, 2019, 67(23): 6039–6052. doi: 10.1109/TSP.2019.2951223.
[6]	TANAKA T and MANDIC D P. Complex empirical mode decomposition[J]. IEEE Signal Processing Letters, 2007, 14(2): 101–104. doi: 10.1109/LSP.2006.882107.
[7]	RILLING G, FLANDRIN P, GONCALVES P, et al. Bivariate empirical mode decomposition[J]. IEEE Signal Processing Letters, 2007, 14(12): 936–939. doi: 10.1109/LSP.2007.904710.
[8]	REHMAN N U and MANDIC D P. Empirical mode decomposition for trivariate signals[J]. IEEE Transactions on Signal Processing, 2010, 58(3): 1059–1068. doi: 10.1109/TSP.2009.2033730.
[9]	REHMAN N and MANDIC D P. Multivariate empirical mode decomposition[J]. Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2010, 466(2117): 1291–1302. doi: 10.1098/rspa.2009.0502.
[10]	ASGHAR M A, KHAN M J, RIZWAN M, et al. AI inspired EEG-based spatial feature selection method using multivariate empirical mode decomposition for emotion classification[J]. Multimedia Systems, 2022, 28(4): 1275–1288. doi: 10.1007/s00530-021-00782-w.
[11]	LANG Xun, ZHANG Yufeng, XIE Lei, et al. Detrending and denoising of industrial oscillation data[J]. IEEE Transactions on Industrial Informatics, 2023, 19(4): 5809–5820. doi: 10.1109/TII.2022.3188844.
[12]	HUANG Guoqing, PENG Liuliu, KAREEM A, et al. Data-driven simulation of multivariate nonstationary winds: A hybrid multivariate empirical mode decomposition and spectral representation method[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2020, 197: 104073. doi: 10.1016/j.jweia.2019.104073.
[13]	LANG Xun, ZHENG Qian, ZHANG Zhiming, et al. Fast multivariate empirical mode decomposition[J]. IEEE Access, 2018, 6: 65521–65538. doi: 10.1109/ACCESS.2018.2877150.
[14]	蔡念, 黄威威, 谢伟, 等. 基于互补自适应噪声的集合经验模式分解算法[J]. 电子与信息学报, 2015, 37(10): 2383–2389. doi: 10.11999/JEIT141632. CAI Nian, HUANG Weiwei, XIE Wei, et al. Ensemble empirical mode decomposition base on complementary adaptive noises[J]. Journal of Electronics & Information Technology, 2015, 37(10): 2383–2389. doi: 10.11999/JEIT141632.
[15]	WANG Yanxue, LIU Fuyun, JIANG Zhansi, et al. Complex variational mode decomposition for signal processing applications[J]. Mechanical Systems and Signal Processing, 2017, 86: 75–85. doi: 10.1016/j.ymssp.2016.09.032.
[16]	ZOSSO D, DRAGOMIRETSKIY K, BERTOZZI A L, et al. Two-dimensional compact variational mode decomposition[J]. Journal of Mathematical Imaging and Vision, 2017, 58(2): 294–320. doi: 10.1007/s10851-017-0710-z.
[17]	孟明, 闫冉, 高云园, 等. 基于多元变分模态分解的脑电多域特征提取方法[J]. 传感技术学报, 2020, 33(6): 853–860. doi: 10.3969/j.issn.1004-1699.2020.06.011. MENG Ming, YAN Ran, GAO Yunyuan, et al. Multi-domain feature extraction of EEG based on multivariate variational mode decomposition[J]. Chinese Journal of Sensors and Actuators, 2020, 33(6): 853–860. doi: 10.3969/j.issn.1004-1699.2020.06.011.
[18]	YAN Xiaoan, LIU Ying, XU Yadong, et al. Multichannel fault diagnosis of wind turbine driving system using multivariate singular spectrum decomposition and improved Kolmogorov complexity[J]. Renewable Energy, 2021, 170: 724–748. doi: 10.1016/j.renene.2021.02.011.
[19]	ZHANG Yijie, ZHANG Haoran, YANG Yang, et al. Seismic random noise separation and attenuation based on MVMD and MSSA[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 5908916. doi: 10.1109/TGRS.2021.3131655.
[20]	CHEN Qiming, CHEN Junghui, LANG Xun, et al. Self-tuning variational mode decomposition[J]. Journal of the Franklin Institute, 2021, 358(15): 7825–7862. doi: 10.1016/j.jfranklin.2021.07.021.
[21]	CHEN Shiqian, YANG Yang, PENG Zhike, et al. Adaptive chirp mode pursuit: Algorithm and applications[J]. Mechanical Systems and Signal Processing, 2019, 116: 566–584. doi: 10.1016/j.ymssp.2018.06.052.
[22]	REHMAN N U and MANDIC D P. Filter bank property of multivariate empirical mode decomposition[J]. IEEE Transactions on Signal Processing, 2011, 59(5): 2421–2426. doi: 10.1109/TSP.2011.2106779.
[23]	LANG Xun, ZHANG Yufeng, XIE Lei, et al. Use of fast multivariate empirical mode decomposition for oscillation monitoring in noisy process plant[J]. Industrial & Engineering Chemistry Research, 2020, 59(25): 11537–11551. doi: 10.1021/acs.iecr.9b06351.
[24]	CHEN Qiming, WEN Qingsong, WU Xialai, et al. Detection and time–frequency analysis of multiple plant-wide oscillations using adaptive multivariate intrinsic chirp component decomposition[J]. Control Engineering Practice, 2023, 141: 105715. doi: 10.1016/j.conengprac.2023.105715.
[25]	YU Wanke, ZHAO Chunhui, and HUANG Biao. MoniNet with concurrent analytics of temporal and spatial information for fault detection in industrial processes[J]. IEEE Transactions on Cybernetics, 2022, 52(8): 8340–8351. doi: 10.1109/TCYB.2021.3050398.
[26]	LINDNER B, AURET L, and BAUER M. A systematic workflow for oscillation diagnosis using transfer entropy[J]. IEEE Transactions on Control Systems Technology, 2020, 28(3): 908–919. doi: 10.1109/TCST.2019.2896223.
[27]	CHEN Qiming, LANG Xun, LU Shan, et al. Detection and root cause analysis of multiple plant-wide oscillations using multivariate nonlinear chirp mode decomposition and multivariate granger causality[J]. Computers & Chemical Engineering, 2021, 147: 107231. doi: 10.1016/j.compchemeng. 2021.107231.
[28]	AFTAB M F, HOVD M, and SIVALINGAM S. Detecting non-linearity induced oscillations via the dyadic filter bank property of multivariate empirical mode decomposition[J]. Journal of Process Control, 2017, 60: 68–81. doi: 10.1016/j.jprocont.2017.08.005.