Show simple item record

Application of T2 control charts and hidden Markov models in predictive maintenance of technical systems

dc.contributor.advisorĐurović, Željko
dc.contributor.otherKovačević, Branko
dc.contributor.otherDenić, Dragan
dc.contributor.otherKvaščev, Goran
dc.contributor.otherTadić, Predrag
dc.creatorKisić, Emilija A.
dc.date.accessioned2016-07-30T15:20:44Z
dc.date.available2016-07-30T15:20:44Z
dc.date.available2020-07-03T08:33:23Z
dc.date.issued2016-06-29
dc.identifier.urihttps://nardus.mpn.gov.rs/handle/123456789/6070
dc.identifier.urihttp://eteze.bg.ac.rs/application/showtheses?thesesId=3451
dc.identifier.urihttps://fedorabg.bg.ac.rs/fedora/get/o:12044/bdef:Content/download
dc.identifier.urihttp://vbs.rs/scripts/cobiss?command=DISPLAY&base=70036&RID=48118799
dc.description.abstractU današnjoj industriji primena najboljih strategija odrţavanja je veoma vaţan zadatak, kako bi se obezbedila stabilnost i pouzdanost tehničkih sistema. U literaturi se moţe naći veliki broj radova i knjiga o različitim strategijama odrţavanja i uglavnom se svuda ističe prednost prediktivnog odrţavanja (eng. predictive maintenance) u odnosu na planirano odrţavanje (eng. time-based maintenance). Prediktivno odrţavanje produţava vreme u kojem sistem radi dobro, smanjuje nepotrebno zaustavljanje sistema, redukuje materijalne gubitke i sprečava nastajanje katastrofalnih otkaza. Iako je sa razvojem visoko sofisticiranih tehnologija ova oblast istraţivanja veoma uznapredovala, i dalje ima puno prostora za poboljšanje postojećih tehnika, kao i za razvoj novih. U ovoj tezi predloţena je inovativna tehnika prediktivnog odrţavanja sa ciljem isticanja prednosti prediktivnog odrţavanja u odnosu na planirano odrţavanje. Predloţeni algoritam primenjen je na konkretan problem koji se javlja planiranim odrţavanjem udarnih ploča mlinova koje melju ugalj u podsistemu za mlevenje uglja u termoelektrani “TEKO”, Kostolac, Srbija. Planiranim odrţavanjem predviĎa se zamena udarnih ploča nakon odreĎenog broja radnih sati, meĎutim u zavisnosti od kvaliteta uglja i same udarne ploče ova zamena je nekad potrebna ranije ili kasnije. Posledice ovakvog odrţavanja su veliki materijalni gubici koji nastaju zbog čestog prevremenog zaustavljanja celog podsistema za mlevenje uglja, kao i mogućnost da otkaz nastupi pre zamene. Prvi korak u primeni predloţenog algoritma bio je sakupljanje podataka, odnosno snimanje akustičkih signala. Akustički signali su snimljeni pomoću usmerenog mikrofona, dok je podsistem za mlevenje uglja u funkciji, bez ometanja rada mlina. Sakupljeni podaci zatim su obraĎeni izdvajanjem korisnih parametara akustičkih signala koji su predstavljeni u frekvencijskom domenu pomoću spektrograma (eng. spectrogram). Inovativnost predloţenog algoritma zasniva se na izboru tehnike prognoze otkaza (eng. failure prognostics). Izabrana je tehnika koja je zasnovana na pokretnim podacima (eng. data driven), zato što se ne traţi poznavanje modela, a dostupni su nam podaci nadgledanja stanja (snimljeni akustički signali). Najpre...sr
dc.description.abstractIn today’s industry, application of the best maintenance strategies is very important task in ensuring stability and reliability of technical systems. Numerous papers and books about different maintenance strategies can be found in literature, and almost everywhere are emphasized merits of predictive maintenance in regard to time-based maintenance. Predictive maintenance extends the period of time during which the system functions well, decreases unnecessary shutdowns, reduces material losses and prevents catastrophic failures. Although this field of research is very much advanced with the development of high sophisticated technologies, there is still a lot of room for improvement of existing techniques and the development of new ones. In this thesis innovative technique of predictive maintenance is proposed with the aim of highlighting the benefits of predictive maintenance compared to time-based maintenance. The proposed algorithm is applied to a specific problem that occurs when time-based maintenance is applied on grinding tables of the coal mill, in coal grinding subsystem at the Thermoelectric Power Plant “TEKO”, Kostolac, Serbia. Time-based maintenance provides replacement of grinding tables after certain number of working hours, but depending on the quality of the coal and grinding table itself, this replacement sometimes needs to be made before or after planned replacement. The consequences of such maintenance are great material losses incurred because of frequent shutdowns of the entire coal grinding subsystem, as well as the possibility that the failure occurs before replacement. First step in application of proposed algorithm was data acquisition, i.e. recording of acoustic signals. Acoustic signals were recorded by means of directional microphone, while the coal grinding subsystem was operating, without interfering with mill operation. Collected data were processed by the feature extraction of acoustic signals that were presented in frequency domain with spectrogram. Innovativeness of the proposed algorithm is based on the selection of failure prognostic technique. Data- driven technique is chosen because model knowledge is not acquired, and condition monitoring data (recorded acoustic signals) are available. Control charts were first applied on extracted parameters of acoustic signals in frequency domain, and then...en
dc.formatapplication/pdf
dc.languagesr
dc.publisherУниверзитет у Београду, Електротехнички факултетsr
dc.rightsopenAccessen
dc.rights.urihttps://creativecommons.org/licenses/by-nc-nd/4.0/
dc.sourceУниверзитет у Београдуsr
dc.subjectPrediktivno održavanjesr
dc.subjectPredictive maintenance. T2 control chart. Hidden Markov Model. Acoustic signalsen
dc.subjectT2 kontrolni dijagramisr
dc.subjectSkriveni Markovljevi modelisr
dc.subjectAkustički signalisr
dc.titlePrimena T2 kontrolnih dijagrama i skrivenih Markovljevih modela na prediktivno održavanje tehničkih sistemasr
dc.titleApplication of T2 control charts and hidden Markov models in predictive maintenance of technical systemsen
dc.typedoctoralThesisen
dc.rights.licenseBY-NC-ND
dcterms.abstractЂуровић, Жељко; Тадић, Предраг; Ковачевић, Бранко; Квашчев, Горан; Денић, Драган; Кисић, Емилија A.; Примена Т2 контролних дијаграма и скривених Марковљевих модела на предиктивно одржавање техничких система; Примена Т2 контролних дијаграма и скривених Марковљевих модела на предиктивно одржавање техничких система;
dc.identifier.fulltexthttps://nardus.mpn.gov.rs/bitstream/id/5027/Disertacija4080.pdf
dc.identifier.fulltexthttps://nardus.mpn.gov.rs/bitstream/id/5028/Kisic_Emilija_Referat_ETF.pdf
dc.identifier.fulltexthttp://nardus.mpn.gov.rs/bitstream/id/5027/Disertacija4080.pdf
dc.identifier.fulltexthttp://nardus.mpn.gov.rs/bitstream/id/5028/Kisic_Emilija_Referat_ETF.pdf
dc.identifier.rcubhttps://hdl.handle.net/21.15107/rcub_nardus_6070


Files in this item

Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record