摘要
针对滚动轴承振动信号的非线性、非平稳性以及复杂性,利用自适应噪声完备集合经验模态分解(Complete ensemble empirical decomposition with adaptive noise,CEEMDAN)的自适应降噪优势,结合多尺度排列熵(Multiscale permutation entropy,MPE)可以反映时间序列在不同尺度上的随机程度以及有效检测到时间序列动力学突变情况等特点,提出一种基于CEEMDAN、多尺度排列熵、Fisher比、GG(Gath-Geva,GG)聚类算法相结合的轴承故障智能识别方法。首先采用CEEMDAN算法对滚动轴承原始信号进行分解,得到若干个含有故障信息的振动信号固有模态函数(Intrinsic mode function,IMF)分量;其次采用峭度准则筛选出一个最优模态分量,并计算其多尺度排列熵值(Multi-scale permutation entropy,MPE);最后,利用Fisher比对MPE特征进行选择,将最终选择的MPE组成特征向量输入到GG聚类模型中,实现滚动轴承故障的智能识别。与其它聚类模型组合方法进行比较的结果证明所提方法在滚动轴承故障识别中的有效性和优越性。
The vibration signals of rolling bearings especially in case of fault occurring are characterized by their nonlinearity,non-stationary and complexity.In this paper,the method of complete ensemble empirical decomposition with adaptive noise(CEEMDAN)is used to solve the problem in virtue of its advantage of adaptive noise reduction.Combined with the multi-scale permutation entropy(MPE),which can reflect the random degree of time series in various scales and effectively detect the sudden dynamic change of the time series,an intelligent bearing fault recognition method is proposed by joint use of CEEMDAN,MPE,Fisher ratio and Gath-Geva(GG)clustering algorithm.Firstly,the original signals of rolling bearings are decomposed by using CEEMDAN and several intrinsic mode functions(IMFs)including fault signals are obtained.Secondly,MPE of the IMF with maximized kurtosis is calculated to obtain an original feature vector.Finally,Fisher ratio is applied to select salient features for feature dimension reduction.The reduced feature vectors are input into GG clustering model to implement intelligent recognition of rolling bearing faults.Comparisons with other clustering-based methods demonstrate the effectiveness and superiority of the proposed method.
作者
熊国良
甄灿壮
张龙
徐天鹏
XIONG Guoliang;ZHEN Canzhuang;ZHANG Long;XU Tianpeng(School of Mechatronics and Vehicle Engineering,East China Jiaotong University,Nanchang 330013,China)
出处
《噪声与振动控制》
CSCD
2020年第6期1-7,28,共8页
Noise and Vibration Control
基金
国家自然科学基金资助项目(51665013,51865010)
江西省自然科学基金资助项目(20161BAB216134,20171BAB206028,20171BAB216030)
江西省研究生创新资金资助项目(YC2018-S248)
江西省研究生创新资金资助项目(YC2019-S243)。
作者简介
熊国良(1962-),男,江西省丰城市人,博士,教授,主要研究方向为复杂动态特性分析与诊断。E-mail:lgxcxx@ecjtu.jx.cn。
