20.500.12556/DKUM-98748 Spektralna razčlenitev tresljajev poškodovanega ležaja s PSoC 6 in UI diplomsko delo Spectral analysis of faulty bearing vibrations using PSoC 6 and AI Naloga obsega teoretični pregled poškodb ležajev in metod za njihovo zaznavo ter praktično izvedbo meritve tresljajev ležajev rabljenega in novega ventilatorja ter presojo prisotnosti poškodb ležajev z metodami z in brez umetne inteligence. Meritve in osnovna obdelava podatkov so bile izvedene z razvojno ploščico PSoC 6 AI Evaluation Board, pri čemer je bil napisan program za tipanje tresljajev z višjo frekvenco tipanja. Za razčlenitev signalov tresljajev sta bili uporabljeni metodi s Fourierjevo transformacijo in razčlenitvijo spektra ovojnice. Zbrani podatki so ob podatkih iz javne zbirke služili tudi za učenje konvolucijske nevronske mreže za klasifikacijo signalov med poškodovane in nepoškodovane. The work encompasses a theoretical review of bearing faults and their detection methods as well as practical execution of bearing vibration measurement of an used and an unused fan with an assessment of presence of faults with or without the use of artificial intelligence. The measurements and basic processing of data is executed with PSoC 6 AI Evaluation Board, where a program was developed for sampling the vibration with a higher sampling frequency. Fourier transform and envelope spectrum analysis methods were used for vibration analysis. The acquired data with a portion of data from a public dataset was used for training a convolutional neural network, meant for classification of signals into those with and those without faults. poškodbe ležajev tresljaji razčlenitev signalov umetna inteligenca nevronske mreže bearing faults vibration signal analysis artificial intelligence neural networks true true false [J. Brglez] Slovenski jezik Angleški jezik Diplomsko delo/naloga 2026-07-05 23:23:29 2026-08-18 10:18:53 2026-09-08 03:08:58 0000-00-00 00:00:00 2026 0 Maribor 0 Maribor 1 spletni vir (1 datoteka PDF (IX, 55 str.)) 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 621.7.08:621.822(043.2) 290197507 UN_Brglez_Jakob_2026.pdf UN_Brglez_Jakob_2026.pdf 1 A7594F6EB639B8E0D08B7D2E2DD06B26 86ef75f90f990df6fba7e50fada5f3b304ab33d07659874f2e6d3382c6f1c057 12d1b2f8-85d5-11f1-a66a-00155d580e34 https://dk.um.si/Dokument.php?lang=slv&id=208517 Fakulteta za elektrotehniko, računalništvo in informatiko 0 0 0