基于RF特征优选和EEMD-SSAE的行星齿轮箱故障诊断

打开文本图片集

中图分类号：TH132.4 文献标志码：A 文章编号：1671-5276（2025）03-0023-05

Fault Diagnosis Planetary Gear Box Based on RF Feature Optimization EEMD-SSAE

LIUWeituan，WANG Youren，JIANGHaoyu

Abstract：Forthelowrecognitionrateinplanetarygearboxfaultdiagnosisduetoinsuficientfeatureextraction，aplanetary gearbox fault diagnosis method combining RFfeatureoptimizationEMD-SSAEis studied.Thetime domain signalis decomposedbyEEMD.IMFcomponentswithlargecorrlationcoeffientsareselectedbasedonPearsoncorelationcoiient， time-domain features，frequencydomainfeaturesoriginal signalfeaturesareextractedtoconstructthedataset.RFis usedto eliminateredundantfeatures，anew dataset isconstructedasinputtothe SSAE network， with stmaxclasifier， faultclassficationisimplemented.Theresultsshowthattheproposedmethodissuperior toother modelsintermsaccuracy robustness under mixed conditions noise interference.

Keywords：faultdiagnosis；planetarygearbox；stackedsparseautoencoder；ensembleempiricalmodedecomposition; feature optimization

0 引言

行星齿轮箱是机械设备中被广泛应用的传动部件。（剩余6974字）