一种物理-虚拟数据融合的数字孪生驱动滚动轴承不平衡故障诊断方法

打开文本图片集

，黄 华（，）

关键词：故障诊断；数字孪生；保真数据；注意力机制；子域自适应机制 中图分类号： 8 文献标志码：A DOI:10.16385/j.cnki.issn.1004-4523.202403018

A method for unbalanced fault diagnosis of rolling bearings driven by digital twin based on physical-virtual data fusion

YANG Pengqiang, HUANG Hua, GUO Baodao, ZHANG Xilong (SchoolofMechanicalandElectricalEngineering，Lanzhou UniversityofTechnology，Lanzhou730o50,China）

Abstract:Inthstudyofunbalancedfaultdanosisofrollingbarings，theraditioaletoditouseaallnumbeofultplesto generatedataadtegeeratedataaediultveheitesuingiowcuacyofsubseqentfultdagnosis.igatthis problem，ametodforunbalancedfultdagnosisofdgita-twin-drivenolingbearingsbasedopsical-virtaldatafusionispoposed Acordingtoieaacactesisfaiialodelollanisabsddelitt dataareobtaiedTefaultfeaturesofrealdataandtwindataareextractedbyusingtheconstructeddeeplearingnetwork.Teaention mechanisminthedeeplearingnetworkhancesthefaultfeaturesandotiizestefeaturemappingTesubdomainadaptivemechanismis introducedtominimizethefeaturedisimilaritybetweenrealdataandtwindata，whichimprovesthediagnosticaccuracyThrough experimentalverificationandcomparaieaalsiswitoermetds，itsprovedtatthemetodproposedintisstudyasteabilityto solve the problem of unbalanced fault diagnosis of rolling bearings.

Keywords:fault diagnosis;digital twin; fidelitydata;atention mechanism;subdomainadaptivemechanism

滚动轴承作为机械传动系统的关键部件，在风力发电机、精密机床和航空发动机等机械装备中被广泛应用1]。（剩余24812字）