31 KB316 KB32 KB199920252005Arenas, Jorge, P.številka:4letnik:519 stranistr. 184-192ISSN:0039-2480COBISSID:8336411URN:URN:NBN:SI:doc-QX0MZ7ONenAssociation of Mechanical Engineers and Technicians of Slovenia et al.Zveza strojnih inženirjev in tehnikov Slovenije et al.Strojniški vestnikanalize šumakotalni ležajiležaji kotalnimerjenjesignali nihanjašumzanke krmilnePovečanje signala nihanja kotalnih ležajev z uporabo prilagodljive krmilne zanke za zmanjševanje šuma| Enhancing the vibration signal from rolling contact bearing using an adaptive closed-loop feedback control for wavelet de-noising|Several techniques in both the time and frequency domains have been reported for the condition monitoring and fault diagnosis of equipment and processes. The monitoring and diagnosis is accomplished through the analysis and interpretation of signals acquired from sensors and transducers. However, any structure-borne vibration propagated through the neighbouring structures will produce a background vibration in which the required vibration signals for thediagnosis are often submerged, in particular during the early stage of failure development. If the background-noise level is too high or when the bearing vibration signature level is too low, traditional techniques such as wavelet de-noising analysis can be ineffective in cancelling the noise of suchsignals. In this paper the combination of an adaptive signal enhancement and the wavelet transform for de-noising a vibration signal measured on a rolling contact bearing is presented. The normalized least mean-square and recursive least-square algorithms were used as the adaptive weight-control mechanism. The final aim of the adaptive filter was to minimize the mean-square value of the error signal, which implies the maximization of the output signal-to-noise ratio of the system. The results showed that a combination of the adaptive vibration signal enhancement and the wavelet transform yielded the best signal-to-noise ratio. This means that the result can reveal hidden signal structures that are directly associated with a bearing's internal defectPredstavljenih je bilo ze več metod s področij časa in frekvence za nadzorovanje razmer in ugotavljanje napak na opremi oz. postopkih. Nadzorovanje in ugotavljanje izvajamo z analizami in tolmačenjem signalov, pridobljenih z zaznavali in pretvorniki. Vendar pa vsako notranje nihanje, ki se prenasa preko sosednjih teles, v ozadju povzroči nihanje, v katerem se signal, ki ga potrebujemo za analize, pogosto izgubi, se posebej v zgodnejši fazi nastanka napake. Če je šumnost nihanja v ozadju prevelika oz. je signal nihanja ležajev premajhen, so lahko običajne metode, kot npr. analiza zmanjševanja šuma valovanja, neučinkovite pri izničenju suma takih signalov. V prispevku smo predstavili kombinacijo poveeanja prilagodljivega signala in spremembo valovanja za zmanjšanje šuma signala nihanja, izmerjenega na kotalnih ležajih. Kot postopek prilagodljivega uteznega krmiljenja smo uporabili algoritma normaliziranega najmanjšega srednjega kvadrata in rekurzivnih najmanjših kvadratov. Končni namen prilagodljivega filtra je bil zmanjšanje srednje kvadratične vrednosti signala napake, kar pomeni povečanje razmerja izstopnega signala in šuma sistema. Rezultati so pokazali, da povečanje prilagodljivega signala nihanja in sprememba valovanja povzročita najboljše razmerje signala in šuma. Kar pomeni, da lahko rezultat odkrije skrite sestave signala, ki so povezane neposredno z notranjimi okvarami v ležajihTEXTznanstveno časopisjejournalsSlovenian National E-content AggregatorNational and University Library of SloveniaUniverza v Ljubljani, Fakulteta za strojništvo