基于改进EEMD-FFT-FRFT的非平稳故障特征提取方法

打开文本图片集

关键词：非平稳信号处理；故障特征提取；故障诊断；改进EEMD-FFT-FRFT法；变转速工况；强噪声环境中图分类号：TN911.7-34；TH113.1 文献标识码：A 文章编号：1004-373X（2026）01-0129-08

Non-stationaryfaultfeatureextractionmethodbasedonimproved EEMD-FFT-FRFTmethod

JIANG Bangyu1²， ZHANG Zhenghua1，²，MENG Da1，2 ，LIBin†²，ABUDUKAYIMUAbulimiti3 （1.CollegeofInformationandArtificial Intelligence，Yangzhou University，Yangzhou 225127，China; 2.JiangsuProvinceEngineeringResearchCenterofKnowledgeManagementandIntelligentService，Yangzhou225127，Cina; 3.Yangzhou Zhishui Internet of ThingsCo.，Ltd.，Yangzhou 2256OO，Chia）

Abstract：Fault feature extractionof non-stationary signals incomplex industrial environment remains chalenging inthe fieldof signalprocessingandfaultdiagnosis，particularlyunderstrongnoiseinterferenceandof-designconditions.Thefeature extractionefectoftheexistingmethodsisstilldificulttomeetheengineeringrequirementsInviewoftheabove，thispaper presentsanenancedsignalprocessingmethodbasedonimprovedfractionalFouriertransform （FRFT）toimprovetheextractionefect offaultfeaturesofstatioaysignals.Firstlyhecolltedvbratiosignalsarepreproessdbynveloedemodulationadan normalization，and then decomposedbyensembleempirical modedecomposition（EEMD）toavoidmodalaliasing inEMD. Secondly，thefastFouriertransform（FF）isutilizedtoobjectivelyselectintrinsicmodefunction（IMF）containingkeyfeatures. Finaly，theFispdfortesletedI，sostaleaturetractdseuppresionExpeimetalation onarollngbearingplatformdemonstratesthat theproposed methodextractsfaulteatures morecompletelyandclearlywithreduced noiseincomparison withtheexisting algorithms.Tosumup，theefectiveness of theproposed method has beenconfirmed.

KeyWords：non-stationary signalprocessing;fault feature extraction;faultdiagnosis;improved EEMD-FFT-FRFTmethod; variable-speed operatingcondition;high-noiseenvironment

0 引言

生产的关键。（剩余8777字）