基于改进VMD的管道声发射信号去噪算法  

Improved VMD-based Noise Reduction Algorithm for Pipeline Acoustic Emission Signals

作　　者：席剑辉 修文斌 刘艳梅 汪炜皎 XI Jianhui;XIU Wenbin;LIU Yanmei;WANG Weijiao(School of Automation,Shenyang Aerospace University)

机构地区：[1]沈阳航空航天大学自动化学院

出　　处：《管道技术与设备》2024年第6期31-36,共6页Pipeline Technique and Equipment

基　　金：国家自然科学基金项目(61203352)。

摘　　要：针对低信噪比管道声发射信号去噪问题,提出一种改进变分模态分解(VMD)的信号去噪算法。首先,引入蚁群优化算法(ACO)对VMD中的2个关键参数(惩罚因子和模态分解数)进行全局寻优,找到有效的含噪本征模态函数IMF;其次,利用小波软阈值函数去噪处理局部弱相关性IMFs,有效保留强噪声背景下的弱有效信息,减小信号误差,同时降低计算量;最后,重构小波阈值处理的IMFs和相关性强的IMFs,得到去噪后的信号。通过仿真和实测信号进行实验验证。实验结果表明,该算法显著提高了信号的信噪比,在管道声发射信号去噪方面展现出良好的性能。An improved signal denoising algorithm based on enhanced variational mode decomposition(VMD)to address the denoising of low signal-to-noise ratio pipeline acoustic emission signals was proposed.Firstly,the ant colony optimization(ACO)was introduced to optimize two key parameters in VMD,the penalty factor,and the number of mode decompositions,thereby finding the effective intrinsic mode function IMF with noise.Secondly,wavelet thresholding function was used to denoise the local weak correlation IMFs,effectively retaining the weak effective information in the background of strong noise,and reducing the signal error and the calculation amount.Finally,the denoised signal was reconstructed by combining the wavelet-thresholder IMFs with strongly correlated IMFs.The experimental results are verified by simulation and measured signals.Experimental results show that the proposed algorithm significantly improves the signal to noise ratio,showing good performance in the pipeline acoustic emission signal denoising.

关 键 词：变分模态分解 小波阈值去噪 蚁群优化算法 相关系数 

分 类 号：TN911[电子电信—通信与信息系统] TP18[电子电信—信息与通信工程]