基于CEEMDAN和数学形态学算子的轴振信号降噪研究  

Denoising Research of Shaft Vibration Signal Based on CEEMDAN and Mathematical Morphology Operator

作　　者：田元兴 苏宝定 于波 于佳鑫 Tian Yuanxing;Su Baoding;Yu Bo;Yu jiaxin(China Guangdong Nuclear Wind Power Co.,Ltd.,Beijing 100070,China;Key Laboratory of Power Station Energy Transfer Conversion and System,Ministry of Education,North China Electric Power University,Beijing 102206,China)

机构地区：[1]中广核风电有限公司,北京100070 [2]华北电力大学电站能量传递转化与系统教育部重点实验室,北京102206

出　　处：《水动力学研究与进展(A辑)》2024年第6期952-959,共8页Chinese Journal of Hydrodynamics

摘　　要：为了解决电站轴振信号噪声问题,该文提出了一种基于完全自适应噪声集合经验模态分解、排列熵以及数学形态学算子的降噪方法CEEMDAN-PE-AVG。针对不同噪声强度的轴振信号,首先使用CEEMDAN将含噪轴振信号分解为若干分量,其次通过排列熵分析,区分有用信号主导分量和噪声主导分量,最后采用数学形态学对有用信号主导分量进行滤波,并将滤波后的信号相加得到降噪后的轴振信号。与基于小波软阈值去噪方法比较,基于本文方法得到的降噪结果信噪比更高,均方根误差更低,互相关系数更高。In order to solve the noise problem of shaft vibration signal of power station,a denoising method CEEMDAN-PE-AVG is proposed in this paper,based on complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN),permutation entropy and mathematical morphology operator.Firstly,for shaft vibration signals with different noise intensity,the noisy shaft vibration signals are decomposed into several components using CEEMDAN.Secondly,components dominated by the useful signals and the components dominated by noise are distinguished by permutation entropy analysis.Then,the components dominated by the useful signals are filtered by a mathematical morphology operator.Finally,the filtered signals are combined to realize the denoising of the shaft vibration signals.By comparing with the denoising method based on wavelet soft threshold denoising,the denoising results based on the method in this paper have a higher signal-to-noise ratio,lower root mean square error,and higher correlation coefficient.

关 键 词：信号分解 排列熵 数学形态学 轴振信号 信号降噪 

分 类 号：TH133.2[机械工程—机械制造及自动化]