作者：李凯丰, 王浩全, 侯清
作者单位：中北大学信息与通信工程学院，山西 太原 030051
关键词：超声信号;信号去噪;小波分析;变分模态分解
摘要：
超声信号在空气中传播效率低、气固界面耦合时透射率差，且易受到接收电路固有噪声和材料晶粒散射噪声的影响，导致接收的信号信噪比低。针对这一问题，结合小波分析“信号显微镜”的优点和变分模态分解(vmd)在窄带谐波信号提取方面的优势，提出一种基于小波分析联合vmd的超声信号去噪方法，并且对小波的阈值做改进，以减少信号的损失。利用改进阈值的小波分析对信号预处理，实现信号和噪声的初步分离，然后用vmd分解信号，提取所需频带的分量，对小波处理的结果进行优化，达到最终去噪的目的。实验数据表明，处理后的超声信号与原始信号相比，snr平均提高1.84 db，mse减小34%，改进阈值后时域峰值损失减少7%，信号能量损失减少13%。该方法去噪彻底，有用信号保留完整，为超声信号的去噪提供新思路。
research on ultrasonic signal denoising method based on wavelet analysis combined with vmd
li kaifeng, wang haoquan, hou qing
school of information and communication engineering, north university of china, taiyuan 030051, china
abstract: the received ultrasonic signal has poor signal to noise ratio due to low propagation efficiency in the air, poor transmittance when the gas-solid interface is coupled, noises affected by the inherent noise of the receiving circuit, and the scattering noise of material grains. to solve this problem, combined with the advantages of wavelet analysis "signal microscope" and the superiority of variational modal decomposition (vmd) in narrowband harmonic signal extraction, an ultrasonic signal denoising method based on wavelet analysis and vmd is proposed. and the threshold of wavelet is improved to reduce the loss of signal. the signal is preprocessed by using the wavelet analysis of the improved threshold to achieve the initial separation of the signal and noise. and then the signal is decomposed with vmd, the components of the required frequency band are extracted. the results of wavelet processing are optimized by vmd to achieve the purpose of denoising. the experimental data show that the snr of the processed ultrasonic signal is increased by 1.84 db on average and the mse is reduced by 34% compared with the original signal. after improving the threshold, the time domain peak loss is reduced by 7% on average, and the signal energy loss is reduced by 13%. this method denoises completely and keeps the useful signal intact, which provides a new idea for the denoising of ultrasonic signals.
keywords: ultrasonic signal;signal denoising;wavelet analysis;variational modal decomposition
2023, 49(4):52-59 收稿日期: 2022-07-11;收到修改稿日期: 2022-10-14
基金项目:
作者简介: 李凯丰(1996-),男,山西长治市人,硕士研究生,专业方向为超声无损检测、信号处理
参考文献
[1] mousavi m, gandomi a h. wood hole-damage detection and classification via contact ultrasonic testing[j]. construction and building materials, 2021, 307: 124999
[2] fang y m, lin l j, feng h l, et al. review of the use of air-coupled ultrasonic technologies for nondestructive testing of wood and wood products[j]. computers and electronics in agriculture, 2017, 137: 79-87
[3] 常俊杰, 吴中权, 罗文斌, 等. 基于eemd-vmd的空气耦合超声储油罐乳化层液位检测[j]. 中国测试, 2022, 48(5): 31-37
[4] 董明, 马宏伟, 陈渊. 基于经验模式分解的超声信号去噪算法[j]. 煤炭技术, 2016, 35(9): 269-270
[5] sharma g k, kumar a, jayakumar t, et al. ensemble empirical mode decomposition based methodology for ultrasonic testing of coarse grain austenitic stainless steels[j]. ultrasonics, 2015, 57: 167-178
[6] 刘奎, 张冬梅, 于光, 等. 空气耦合超声信号的小波阈值滤噪试验研究[j]. 机械工程学报, 2015, 51(20): 61-66
[7] 张曼, 王明泉, 杨顺民, 等. 基于小波分析的锂电池空耦超声信号降噪研究[j]. 测试技术学报, 2020, 34(4): 304-310
[8] sharma n k, samantaray s r, bhende c n. vmd-enabled current-based fast fault detection scheme for dc microgrid[j]. ieee systems journal, 2021, 16(1): 933-944
[9] 董鑫, 李国龙, 何坤, 等. 谱图小波阈值降噪及其在滚刀主轴振动信号分析中的应用[j]. 机械工程学报, 2020, 56(11): 96-107
[10] dragomiretskiy k, zosso d. variational mode decomposition[j]. ieee transactions on signal processing, 2014, 62(3): 531-544
[11] 温益凯, 陈乐, 富雅琼. 基于小波降噪的蜂鸣声快速识别方法[j]. 中国测试, 2022, 48(4): 12-17 34
[12] bi f r, li x, liu c c, et al. knock detection based on the optimized variational mode decomposition[j]. measurement, 2019, 140: 1-13
[13] 郭佳辉, 高一轩, 刘昌伟, 等. 基于vmd变分模态分解的冲击波信号时频分析[j]. 弹箭与制导学报, 2022, 42(1): 118-122
[14] 肖洒, 陈波, 沈道贤, 等. 改进vmd和阈值算法在局部放电去噪中的应用[j]. 电子测量与仪器学报, 2021, 35(11): 206-214
