摘要
为了克服傅里叶变换、经验模态分解与傅里叶分解方法在分析非平稳信号方面的不足,提出一种适合非线性和非平稳信号分析的新方法——自适应经验傅里叶分解(Adaptive empirical Fourier decomposition,AEFD)。AEFD方法以快速傅里叶变换为基础,通过对变换系数进行分组重构,能够将一个非平稳信号自适应地分解为若干个瞬时频率具有物理意义的傅里叶本征模态函数(Fourier intrinsic mode function,FIMF)之和。研究了AEFD的分解正交性和精确性,通过仿真信号分析,将其与经验模态分解,变分模态分解和傅里叶分解方法等进行了详细对比,结果表明了AEFD的优越性。最后,为了提高故障诊断的精度和验证AEFD的有效性,将AEFD应用到转子碰摩和滚动轴承局部故障诊断中。试验数据分析结果表明,与经验模态分解等方法相比,AEFD不仅能够有效地诊断故障,而且诊断精度更高。
A novel non-stationary signal analysis method termed adaptive empirical Fourier decomposition(AEFD)is proposed to overcome the deficiencies of Fourier transform,empirical mode decomposition(EMD)and Fourier decomposition method(FDM)in non-stationary signal analysis.AEFD is based on fast Fourier transform and by grouping and reconstructing the coefficients of fast Fourier transform,it can adaptively decompose a given non-stationary signal into several Fourier intrinsic mode functions(FIMF)with instantaneous frequency of physical significance.The decomposition orthogonality and accuracy of AEFD are also studied.The proposed method is compared with EMD,FDM and variational mode decomposition in detail through simulation signal analysis and the results have verified the superiority of AEFD.Finally,AEFD method is applied to the diagnosis of rotor system with local rubbing and rolling bearing with local fault to verify its effectiveness and improve the accuracy of fault diagnosis.The experimental data analysis results show that compared with EMD,AEFD can effectively identify the fault location and get a higher diagnostic accuracy than the methods mentioned above.
作者
郑近德
潘海洋
程军圣
包家汉
刘庆运
丁克勤
ZHENG Jinde;PAN Haiyang;CHENG Junsheng;BAO Jiahan;LIU Qingyun;DING Keqin(School of Mechanical Engineering,Anhui University of Technology,Maanshan 243032;School of Mechanical and Vehicle Engineering,Hunan University,Changsha 410082;China Special Equipment Inspection and Research Institute,Beijing 100029)
出处
《机械工程学报》
EI
CAS
CSCD
北大核心
2020年第9期125-136,共12页
Journal of Mechanical Engineering
基金
国家重点研发计划(2017YFC0805100)
国家自然科学基金(51975004)
安徽省高校自然科学研究项目(KJ2019A053,KJ2018ZD005)资助项目。
关键词
非平稳信号
经验模态分解
变分模态分解
自适应经验傅里叶分解
故障诊断
non-stationary signal
empirical mode decomposition
variational mode decomposition
adaptive empirical Fourier decomposition
fault diagnosis