資源描述:
《基于故障演化分析的齒輪故障預(yù)測特征選擇-論文.pdf》由會員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在行業(yè)資料-天天文庫。
1、第32卷第4期深圳大學(xué)學(xué)報理工版Vo1.32No.42015年7月JOURNALOFSHENZHENUNIVERSrISCIENCEANDENGINEERINGJuly2015【電子與信息科學(xué)/ElectricsandElectronicsInformation】基于故障演化分析的齒輪故障預(yù)測特征選擇傅鵬程,王昌明,段俊斌,譚曉棟1)南京理工大學(xué)機(jī)械工程學(xué)院,南京210094;2)北京特種機(jī)電技術(shù)研究所,北京100012;3)武警警官學(xué)院電子技術(shù)系,成都610213摘要:提出基于故障演化分析的齒輪故障預(yù)測特征選擇方法.根據(jù)機(jī)械動力學(xué)與振動模型。
2、計(jì)算不同故障嚴(yán)重程度下系統(tǒng)的輸出響應(yīng).使用特征選擇方法建立不同特征描述的齒輪故障演化曲線,計(jì)算各特征對齒輪故障演化過程的跟蹤能力,選擇跟蹤能力最大的特征作為故障預(yù)測特征.以一階齒輪傳動系統(tǒng)的齒輪裂紋故障為例,論證結(jié)果表明,小波熵對齒輪裂紋故障演化過程的跟蹤能力最大,采用小波熵作為齒輪裂紋故障的預(yù)測特征能有效提高故障預(yù)測的精度.關(guān)鍵詞:機(jī)械動力學(xué)與振動;齒輪;特征選擇;故障預(yù)測;故障演化分析;跟蹤能力;小波熵中圖分類號:TP165文獻(xiàn)標(biāo)志碼:Adoi:10.3724/SP.J.1249.2015.04434Afeatureselectionfor
3、faultprognosisbasedonfaultevolutionanalysisFuPengcheng’甜,WangChangming,DuanJunbin,andTanXiaodong1)SchoolofMechanicalEngineering,NamingUniversityofScienceandTechnology,Nanjing210094,P.R.China2)BeijingSpecialElectromechanicalTechnologyResearchInstitute,Beijing100012,P.R.China3
4、)DepartmentofElectronicTechnology,OfficersCollegeofCAPF,Chengdu610213,P.R.ChinaAbstract:Inordertoimprovethereliabilityandtheeffectivenessofgearfaultprognosis,weproposeafeatureselectionbasedonfaultevolutionanalysis.Firstly,webuildamechanicaldynamicsandvibrationmodel,whichacco
5、rdinglygeneratesdynamicresponseswithtoothdamagegrowth.Secondly,weobtainthetrendcurvesofgearfaultevolutionforseveralfeatures,andcalculatethetrackabilityforgearfaultevolutionofeachfeature.Thefeaturewiththestrongesttrackabilityischosenasthefeatureforgearfaultprognosis.Finally,w
6、euseagearcrackfaultinaone—stagegearboxtoverifytheperformanceoftheproposedmethod.Resultsshowthatthewaveletenergyentropywiththeoptimaltrackabilityforgearcrackevolutionamongseveralselectedfeaturesisthemostsuitablefeatureforgearcrackfaultprognosis.Keywords:mechanicaldynamicsandv
7、ibration;gear;featureselection;faultprognosis;faultevolutionanalysis;trackability;waveletentropy故障預(yù)測技術(shù)對有效避免處于工作狀態(tài)的部件義.故障預(yù)測特征選擇方法是實(shí)現(xiàn)故障預(yù)測功能因突然失效而導(dǎo)致災(zāi)難性事故的發(fā)生具有重要意的關(guān)鍵,它決定了故障預(yù)測的準(zhǔn)確和有效.目Received:2015—01.25;Revised:2015-05—22;Accepted:2015—06—15Foundation:TheSpecializedResearchFundfo
8、rtheDoctoralProgramofHigherEducation(20133219110027)十Correspondingauthor:Se