常州大学学报（自然科学版）2025，Vol.37Issue(3)：75-86,92,13.DOI:10.3969/j.issn.2095-0411.2025.03.009
基于CELMDAN与IMOMEDA的微弱机械特征增强方法
Weak mechanical feature enhancement method combining CELMDAN with IMOMEDA
顾张清 1黄清岩 1任世锦 1郝国生1
作者信息
- 1. 江苏师范大学计算机科学与技术学院,江苏徐州 221116
- 折叠
摘要
Abstract
To address the issue of equipment background noise degrading mechanical fault detection,a novel weak mechanical feature enhancement method was presented by combining complete ensemble local mean decomposition with adaptive noise(CELMDAN)with improved multipoint optimal mini-mum entropy deconvolution adjusted(IMOMEDA).Firstly,CELMDAN was utilized to decompose complex vibration signals into multiple single modal product functions(PFs),dealing with the prob-lem of ensemble local mean decomposition(ELMD)difficult determination of noise amplitude imposed to signals and trial times.Secondly,a periodic modulation intensity(PMI)with strong robustness,clear physical meaning and scale invariance was proposed to screen out the effective PFs;then,the IMOMEDA method was proposed to eliminate the noise in the selected PFs.This method can adap-tively extract periodic fault transient features from vibration signals by iteratively estimating the opti-mal model parameters,and the spectral kurtosis of transients in the frequency domain can be found.As a result,weak fault features submerged by background noise are effectively enhanced.The incipi-ent fault related impact components can be the optimal IMOMEDA parameters.As a result,enhance-ment can be yielded in an iterative way.This method can locate the spectral kurtosis of transients in the frequency domain,thereby extracting weak fault features submerged by background noise.Final-ly,taking the coal mine hoist as the research object,we designed a variety of vibration signal feature enhancement method comparison experiments,mechanical operation state diagnosis performance ex-periments and signal feature enhancement algorithm comparison experiments,which verified the va-lidity of this paper's method from multiple perspectives.关键词
微弱故障信号增强/改进的多点最优最小熵去卷积调整/自适应噪声完全集成局部均值分解/周期调制强度/煤矿提升机Key words
weak mechanical signal enhancement/improved multipoint optimal minimum entropy de-convolution adjusted/complete ensemble local mean decomposition with adaptive noise/periodic modulation intensity/mine hoist分类
机械工程引用本文复制引用
顾张清,黄清岩,任世锦,郝国生..基于CELMDAN与IMOMEDA的微弱机械特征增强方法[J].常州大学学报（自然科学版）,2025,37(3):75-86,92,13.
