面向不同标签与域配置的统一跨域故障诊断方法

打开文本图片集

关键词：智能故障诊断；多场景跨域诊断；统一方法；迁移学习 中图分类号： TH165+.3 文献标志码：A DOI：10.16385/j.cnki.issn.1004-4523.202508021

Unified cross-domain fault diagnosis method towards different label and domain configurations

ZHANG Yuteng'，LYU Yufan'， KONG Yun1，2，3 ，CHEN Ke1，4 ， YAN Zhiwu4， DONG Mingming'， CHU Fulei5 （1.School of Mechanical Engineering，Beijing Institute of Technology，Beijing 10oo81，China; 2.State KeyLaboratoryofMechanicalTransmisionforAdvancedEquipment，Chongqing University，Chongqing 444，China; 3.TangshanResearchInstitute，BeijingIstituteofTechnology，TangshanO6315，China；4InnerMongoliaFirstMachneryGroupCo， Ltd.，Baotou014032，China；5.Departmentof Mechanical Enginering，Tsinghua University，Beijing 10084，China）

Abstract：Reliablefaultdagosisisucialtoesuingthesafeandeficientoperatioofigndindustrialuipent.Co-i inteligentdagnosistechnologiesbasedonusuperviseddomainadaptation（UDA）havedemonstratedpromisingapicationprospctsn scenariossuchaso-eqpmentandvaableorkingtrasferdiagnosisonditos.However，eirfectiveesshiglyeliesonpecific priorassmptiosregardingteinter-domainlabelelatiosipsanddomainonfiguratios，ichlargelyestritstegeneralizabilityad practicalityofUtchqusiactalidustrialfultdiagosisseariosstebeissstspapepropesdomainfaultdagnosisfmeorkaplicabletodiferentlbelanddomaiofiguratiosTeproposedframekcostructsapredictielass confusio（PC）biassharedacrossmultiplesenariostoguidecross-domainknowledgetransfer，enablingadaptationtovarioustransfer diagnosticscenariosToacuratelymeasurethetendencyofthePCCbias，aprototypesimilaritybasedfaultdiscriminationmethodis developed，whichenancesclasificationrobustessandprovidesreliablepredictiondistributionstoestimatetheCCbias.Tenalabel smothing-basedprobabilitycalibrationmethodisdsigedforprobabilityregularzationallvatingtheunderestimationofteCCbias causedbyovercofidentpredictionExpermetalvalidatioesultsoaplanetarygarboxtrasmisiosystematasetdmonstratethathe proposed method achieves an average diagnostic accuracy of 98.37% across four cross-domain diagnostic scenarios with different label and domainconfigurations，outperformingstate-of-the-artapproachesandfullyverifyingitsgeneralityandsuperiority

Keywords：intelligent fault diagnosis;multi-scenario cross-domain diagnosis;unified method; transfer learning

在现代工业体系中，航空发动机、风力发电机、电力机车等高端装备是航空航天、能源交通等关键领域的核心，其运行状态直接关乎工业生产的安全性与稳定性。（剩余19773字）