一種基于DCNN-LSTM混合模型的RUL預(yù)測(cè)方法_第1頁(yè)
一種基于DCNN-LSTM混合模型的RUL預(yù)測(cè)方法_第2頁(yè)
一種基于DCNN-LSTM混合模型的RUL預(yù)測(cè)方法_第3頁(yè)
全文預(yù)覽已結(jié)束

下載本文檔

版權(quán)說(shuō)明：本文檔由用戶提供并上傳，收益歸屬內(nèi)容提供方，若內(nèi)容存在侵權(quán)，請(qǐng)進(jìn)行舉報(bào)或認(rèn)領(lǐng)

文檔簡(jiǎn)介

一種基于DCNN-LSTM混合模型的RUL預(yù)測(cè)方法AbstractRemainingUsefulLife(RUL)predictionplaysavitalandsignificantroleinensuringthesafetyandreliabilityofsystems,especiallyintheaerospace,energy,andtransportationindustry.Inthispaper,weproposeaHybridmodelcombiningconvolutionneuralnetwork(CNN)andLongShort-TermMemory(LSTM)forRULprediction.Theproposedmodelextractsthefeatureinformationofthesensordatabyaseriesofconvolutionneuralnetworklayers,thenutilizestheLSTMnetworktolearnthetemporaldependenciesbetweenthefeatureinformation.WeevaluatetheproposedmethodontheCMAPSSdataset,andtheresultsindicatethesuperiorityandeffectivenessoftheproposedmethod.IntroductionRemainingUsefulLife(RUL)predictionisacrucialtaskforensuringthesafetyandreliabilityofsystems.AccurateRULpredictionprovidesthemeanstoperformpreventativemaintenanceattheoptimaltime,hencereducingequipmentdowntime,maintenancecosts,andenergyconsumption.ManyRULpredictionmethodshavebeenproposed,includingmodel-basedmethods,data-drivenmethods,andhybridmethods.Amongthem,data-drivenmethodshaveattractedgreatattentioninrecentyearsduetoitshighaccuracy,robustness,andsimplicity.Convolutionalneuralnetworks(CNN)andLongShort-TermMemory(LSTM)networksarepopulardeeplearningmodelsthathavebeenappliedinavarietyoffields,includingsignalprocessing,imagerecognition,naturallanguageprocessing,andothers.Inthispaper,weproposeahybridmodelcombiningCNNandLSTMforRULprediction.MethodologyTheproposedmethodconsistsoftwostages,asshowninFigure1.Inthefirststage,wepreprocessthefeatureinformationofthesensordataandextractthefeaturesthroughaseriesofconvolutionneuralnetworklayers.Inthesecondstage,weutilizetheLSTMnetworktolearnthetemporaldependenciesbetweenfeatureinformation.![image.png](attachment:image.png)Figure1.Architectureoftheproposedmethod.1.CNNFeatureExtractionThehigh-dimensionalsensordataobtainedfromthesystemcontainscomplexandrichinformation,weuseconvolutionneuralnetworklayerstoextracttheessentialfeaturesforRULprediction.TheCNNiscomposedofaseriesofconvolutionallayers(Conv)andpoolinglayers(Pool).EachconvolutionallayerisfollowedbyaReLUactivationfunctionandabatchnormalizationlayer.Thepoolinglayerreducesthedimensionalityandpreservestheessentialfeatures.Theoutputofthelastpoolinglayerisflattenedandfedintoafullyconnectedlayer(FC).Thedropouttechniqueisappliedtopreventoverfitting.TheoutputoftheFClayerisafeaturevectorrepresentingtheessentialcharacteristicsofsensordata.2.LSTMNetworkTheextractedfeatureinformationfromthesensordataisfedintotheLSTMnetworkforlearningthetemporaldependencies.TheLSTMarchitecturecanlearnthelong-termdependenciesinthetimeseriesdatabymaintainingamemorycellandvariousgatesthatcontroltheflowofinformation.TheLSTMnetworkhasthreegates:inputgate(i),outputgate(o),andforgetgate(f).Thesegatescontroltheflowofinformationthroughthememorycell.ThefinaloutputoftheLSTMnetworkisfedintoafullyconnectedlayer(FC)withasigmoidactivationfunctionforRULprediction.DatasetandEvaluationToevaluatetheeffectivenessoftheproposedmethod,weusedtheNASAC-MAPSSdataset,whichiscommonlyusedforRULpredictionbenchmarks.Thedatasetconsistsoffoursubsets,eachcontainingsensormeasurementsfromturbofanengines.ThetaskistopredicttheRULoftheenginebasedonthesesensormeasurements.Foreachsubset,werandomlyselectedatrainingsetof70%ofthedataandkept30%asatestset.Wetrainedtheproposedhybridmodelonthetrainingsetandevaluateditonthetestset.Theperformanceoftheproposedmethodiscomparedwithotherstate-of-the-artRULpredictionmethods,includingSupportVectorMachines(SVM),RandomForest(RF),andGradientBoostingDecisionTree(GBDT).TheevaluationcriteriausedinthispaperareRootMeanSquareError(RMSE),MeanAbsoluteError(MAE),andCoefficientofDetermination(R2).![image-2.png](attachment:image-2.png)Table1.ComparisonofRMSE,MAE,andR2ofdifferentRULpredictionmethodsontheCMAPSSdataset.ExperimentalresultsandanalysisTable1showsthecomparisonofRMSE,MAE,andR2ofdifferentRULpredictionmethodsontheCMAPSSdataset.ItcanbeseenthatourproposedhybridmodelachievesthelowestRMSEandMAE,indicatingthatithasbetterperformanceinpredictingRUL.Moreover,thehybridmodelproducesthehighestR2coefficient,whichisameasureoftheproportionofvarianceintheRULpredictionthatisexplainedbythemodel.WealsovisualizedthepredictedRULandactualRULforthemodelwiththehighestperformanceineachsubset,asshowninFigure2.WecanseethatthepredictedRULishighlyconsistentwiththeactualRUL,indicatingtheexcellentperformanceoftheproposedmethod.![image-3.png](attachment:image-3.png)Figure2.ComparisonofpredictedRULandactualRULontheCMAPSSdataset.ConclusionThispaperproposesahybridmodelcombiningCNNandLSTMforRULprediction.TheproposedmodelextractsthefeatureinformationthroughaseriesofconvolutionallayersandlearnsthetemporaldependenciesbetweenfeatureinformationusingtheLSTMnetwork.TheexperimentalresultsontheCMAPSSdatasetshowthattheproposedmethodhasbetterperformance

溫馨提示

  • 1. 本站所有資源如無(wú)特殊說(shuō)明，都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請(qǐng)下載最新的WinRAR軟件解壓。
  • 2. 本站的文檔不包含任何第三方提供的附件圖紙等，如果需要附件，請(qǐng)聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
  • 3. 本站RAR壓縮包中若帶圖紙，網(wǎng)頁(yè)內(nèi)容里面會(huì)有圖紙預(yù)覽，若沒(méi)有圖紙預(yù)覽就沒(méi)有圖紙。
  • 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
  • 5. 人人文庫(kù)網(wǎng)僅提供信息存儲(chǔ)空間，僅對(duì)用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理，對(duì)用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯，并不能對(duì)任何下載內(nèi)容負(fù)責(zé)。
  • 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容，請(qǐng)與我們聯(lián)系，我們立即糾正。
  • 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時(shí)也不承擔(dān)用戶因使用這些下載資源對(duì)自己和他人造成任何形式的傷害或損失。

最新文檔

評(píng)論

0/150

提交評(píng)論