基于CEEMDAN与多步评估策略的信号降噪与特征增强关键技术研究  

Research on Key Technologies of Signal Denoising and Feature Enhancement Based on CEEMDAN and Multistep Evaluation Strategy

作　　者：苏振华 高林[1] 王少江 刘胜涛 柴星 胡阳虎[2] 邱立朋[2] Su Zhenhua;Gao Lin;Wang Shaojiang;Liu Shengtao;Chai Xing;Hu Yanghu;Qiu Lipeng

机构地区：[1]国机重型装备集团股份有限公司,四川德阳618000 [2]中国重型机械研究院股份公司,陕西西安710018

出　　处：《中国重型装备》2023年第3期64-70,共7页CHINA HEAVY EQUIPMENT

基　　金：四川省重大科技专项(2022ZDZX0040)。

摘　　要：主要阐述了旋转机械在运行过程中,其监测信号存在诸如窄带脉冲、随机噪声等干扰,影响了监测信号对有效特征的表达。为此,提出一种基于CEEMDAN与多步评估策略的信号降噪与特征增强方法。原始信号经过以能量标准差改进的CEEMDAN分解为一系列关键模态,利用相关性分析剔除属于噪声域的模态,进一步地伪模态通过峭度分析被寻找出来,利用基于小波阈值策略实现降噪信号的重构。The interference of monitoring signals such as narrowband pulses and random noise during the operation of rotating machinery,which affects the expression of effective features in the monitoring signals,is described.To this end,a signal denoising and feature enhancement method based on CEEMDAN and multi-step evaluation strategy is proposed.The original signal is decomposed into a series of key modes using CEEMDAN improved by energy standard deviation.Correlation analysis is made to eliminate the modes belonging to the noise domain,and further,the pseudo modes are identified through kurtosis analysis.The denoised signal is reconstructed based on a wavelet threshold strategy.

关 键 词：旋转机械 CEEMDAN 振动信号 多步评估策略 

分 类 号：TH-39[机械工程]