脊提取联合ACMD的变转速滚动轴承故障诊断  

Fault Diagnosis of Rolling Bearings under Variable Speed Based on Ridge Extraction Combined with ACMD

作　　者：李燕文 马萍[1] 王聪[1] 梁城 张浩然 张宏立[1] LI Yanwen;MA Ping;WANG Cong;LIANG Cheng;ZHANG Haoran;ZHANG Hongli(School of Electrical Engineering,Xinjiang University,Urumqi 830000,China)

机构地区：[1]新疆大学电气工程学院,乌鲁木齐830000

出　　处：《噪声与振动控制》2025年第1期89-96,共8页Noise and Vibration Control

基　　金：新疆维吾尔自治区自然科学青年基金资助项目(2022D01C89);国家自然科学基金资助项目(52267010,52065064,62263030)。

摘　　要：变转速工况下滚动轴承故障振动信号受转速波动影响,故障特征易受到噪声和其他无关分量干扰导致时频面模糊,故障特征提取困难。自适应调频模态分解(Adaptive Chirp Mode Decomposition,ACMD)作为新提出的非平稳信号处理算法,可有效分析时变非平稳信号。ACMD需基于希尔伯特变换确定瞬时频率(Instantaneous Frequencies,IF),提取IF的准确性对最终分析结果具有较大影响。因此,在ACMD的基础上,引入脊提取理论,提出基于多时频曲线提取算法(Multiple Time-Frequency Curve Extraction,MTFCE)提取IF的多时频自适应调频模态分解(Multiple Time Frequency ACMD,MACMD)方法。首先对原始振动信号进行包络处理,并通过MTFCE提取其包络图中的IF作为预设频率输入到ACMD算法,然后对包络信号进行ACMD分解,最后根据分解得到的各个信号分量的IF和瞬时幅值(Instantaneous Amplitude,IA)信息构建高分辨率的时频表示,以实现时变非平稳信号的分析。通过分析模拟信号和实测变转速下滚动轴承故障信号可知,该方法能有效诊断时变转速下滚动轴承故障,减少噪声干扰,且具有较强的鲁棒性。Under variable speed conditions,rolling bearing fault vibration signals are affected by speed fluctuations,and fault features are easily interfered by noise and other irrelevant components,resulting in blurred time-frequency planes,making the difficulty to extract fault features.Adaptive Chirp Mode Decomposition(ACMD),as a newly proposed non-stationary signal processing algorithm,can effectively analyze time-varying non-stationary signals.ACMD needs to determine the instantaneous frequency(IF)based on the Hilbert transform,and the accuracy of the extraction of the instantaneous frequency has a great influence on the final analysis results.Therefore,on the basis of ACMD,this paper introduced ridge extraction theory,and proposed Multiple Time-Frequency ACMD(MACMD)method based on Multiple Time-Frequency Curve Extraction(MTFCE)to extract IF.Firstly,the original vibration signal was enveloped,and the instantaneous frequency in its envelope map was extracted by MTFCE as a preset frequency and input to the ACMD algorithm.Then,the envelope signal was decomposed by ACMD.Finally,a high-resolution time-frequency representation was constructed according to the IF and instantaneous amplitude(IA)information of each signal component obtained by the decomposition,so as to realize the analysis of time-varying non-stationary signals.By analyzing the analog signal and the measured fault diagnosis signal of the rolling bearing under variable speed,it was verified that the proposed method can effectively diagnose the rolling bearing fault under time-varying speed,reduce noise interference,and has strong robustness.

关 键 词：故障诊断 变转速 滚动轴承 脊提取 多时频自适应调频模态分解 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置] TN911.6[自动化与计算机技术—控制科学与工程]