資源描述:
《基于KPCA-MVU的噪聲非線性過程故障檢測(cè)方法.pdf》由會(huì)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(kù)。
1、第35卷第12期儀器儀表學(xué)報(bào)Vb1.35NO.122014年12月ChineseJourna1ofScientificInstrumentDec.2014基于KPCA.MVU的噪聲非線性過程故障檢測(cè)方法水陳如清(嘉興學(xué)院機(jī)電工程學(xué)院嘉興314001)摘要:實(shí)際化工過程監(jiān)控?cái)?shù)據(jù)具有非線性特征且易受隨機(jī)噪聲影響。將核主元分析(KPCA)方法與最大方差展開(MVU)特征提取算法相結(jié)合,提出一種基于KPCA.MVU的噪聲環(huán)境下非線性過程故障檢測(cè)新方法。改進(jìn)算法在對(duì)非線性噪聲數(shù)據(jù)的降維過程中,首先對(duì)各樣本點(diǎn)的鄰域范圍采用局部KPCA方法識(shí)別并剔除
2、過程數(shù)據(jù)的噪聲,再提取輸入數(shù)據(jù)空間中的非線性主元;其次,在保持近鄰點(diǎn)間歐式距離不變的前提下,MVU通過旋轉(zhuǎn)平移等變換在低維特征空間中展開高維數(shù)據(jù)流形的同時(shí)保持?jǐn)?shù)據(jù)的全局幾何結(jié)構(gòu)。噪聲環(huán)境下TE過程的仿真分析和丙烯腈聚合過程的實(shí)驗(yàn)研究結(jié)果表明,基于改進(jìn)方法構(gòu)建的過程故障檢測(cè)模型可有效改善基本Mvu和KPCA方法對(duì)非線性噪聲數(shù)據(jù)的特征提取性能,有效增強(qiáng)了對(duì)噪聲的魯棒性。關(guān)鍵詞:最大方差展開;核主元分析;非線性噪聲數(shù)據(jù);故障檢測(cè)中圖分類號(hào):TP27"~TH865文獻(xiàn)標(biāo)識(shí)碼:A國(guó)家標(biāo)準(zhǔn)學(xué)科分類代碼:120.5010Nonlinearproces
3、sfaultdetectionmethodundernoiseenvironmentusingKPCAandMVUChenRuqing(CollegeofMechanicalandElectricalEngineering,JiaxingUniversity,Jiaxing314001,China)ABSTRACT:Actualchemicalprocessmonitoringdatahaves~ongnonlinearbehaviorandareeasilydisturbedbyrandomnoises.Anovelkernelpri
4、nciplecomponentanalysis(KPCA)一maximumvarianceunfolding(MVU)basedfaultdetectionmethodfornonlinearprocessundernoiseenvironmentisproposedbycombiningKPCAandMVUfeatureextractionalgo-rithms.Inthedimensionreductionprocessofnonlinearnoisydata,localKPCAmethodisappliedtoidentifyan
5、deliminatethenoiseintheprocessdataintheneighborhoodofsamplepoints;andthenthenonlinearprincipalcomponentsareex—tractedintheinputdataspace.Next,undertheconditionofkeepingtheEuclideandistancesbetweenneighborpointsun—changed.MVUisUSedtomaptheoriginalhighdimensiondataspacetoa
6、lowdimensionembeddingspacewhilepreservingthedataglobalgeorrtetricstructureviacoordinaterotationandtranslationtransformation.SimulationresultsofTEprocessundernoiseenvironmentandexperimentresultsofacrylonitrilepolymerizationprocessshowthattheimprovedKPCA··MVUbasedfaultdete
7、ctionmodelcanimprovethefeatureextractionperformanceofstandardKPCAandMVUa1.gorithmsfornonlinearnoisydata,andeffectivelyenhancetherobustnessagainstnoise.Keywords:MVU;KPCA;nonlinearnoisydata;faultdetection收稿日期:2014.04ReceivedDate:2014-04·基金項(xiàng)目:浙江省自然科學(xué)基金資助項(xiàng)目(LQ12F03007)2674儀器
8、儀表學(xué)報(bào)第35卷降低數(shù)據(jù)維數(shù),通過MVU算法將非線性流形展開至1引言低維空間并同時(shí)保持其全局結(jié)構(gòu)特性。在特征空間構(gòu)造過程故障監(jiān)測(cè)統(tǒng)計(jì)量,實(shí)現(xiàn)噪聲環(huán)境下非線性過程現(xiàn)代工業(yè)技術(shù)的發(fā)展使過程工業(yè)系統(tǒng)結(jié)構(gòu)日趨的故障