摘要
时变工况下旋转机械的振动信号具有明显的时变调制的特点,熵值方法在提取该类信号特征时具有独特的优势。为了克服传统的熵值方法计算速度慢、熵值不稳定等问题,提出了一种基于精细复合多尺度散度熵的时变工况下旋转机械故障诊断方法,能够更有效地提取故障特征信息并提高故障诊断准确率。首先,采用重采样的方法将时域信号转为角域信号,并利用变分模态分解和独立分量分析相结合的方法对角域信号进行去噪。其次,采用精细复合多尺度散度熵对去噪后的角域信号进行特征提取,然后将提取到的特征输入LR(logistic regression)分类器中识别故障类型。最后,通过时变工况下的齿轮试验对所提方法进行验证,结果表明,所提出的方法有效提高了时变工况下故障诊断准确率。
Vibration signals of rotating machinery under time-varying working conditions have obvious time-varying modulation features,entropy value method has unique advantages in extracting such signals' features.Here,to solve problems of slow calculation speed and unstable entropy value in traditional entropy value method,a fault diagnosis method for rotating machinery under time-varying working conditions based on fine composite multi-scale divergence entropy(FCMDE) was proposed to more effectively extract fault feature information and improve fault diagnosis accuracy.Firstly,a resampling method was used to convert a time-domain signal into an angular domain signal,and a combination of variational mode decomposition and independent component analysis was used to denoise angular domain signals.Secondly,FCMDE was used to extract features from denoised angular domain signals,and the extracted features were then input into a logistic regression(LR) classifier to identify fault types.Finally,the proposed method was verified through gear experiments under time-varying working conditions,and the results showed that the proposed method can effectively improve the accuracy of fault diagnosis under time-varying working conditions.
作者
卢太武
马洪波
王先芝
陈改革
LU Taiwu;MA Hongbo;WANG Xianzhi;CHEN Gaige(School of Communications and Information Engineering(School of Artificial Intelligence),Xi’an University of Posts and Telecommunications,Xi’an 710121,China;School of Automation,Xi’an University of Posts and Telecommunications,Xi’an 710121,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2023年第21期211-218,共8页
Journal of Vibration and Shock
基金
国家自然科学基金(51905399)
陕西省教育厅科技计划(22JK0569)。
关键词
故障诊断
时变工况
精细复合多尺度散度熵
变分模态分解
独立分量分析
fault diagnosis
time-varying working condition
fine composite multiscale divergence entropy(FCMDE)
variational mode decomposition(VMD)
independent component analysis(ICA)
