diplomsko delo
Luka Štajdohar (Author), Janez Bešter (Mentor), Luka Mali (Co-mentor)

Abstract

Dandanes imamo po svetu vedno več senzorjev ter pametnih robnih naprav zaradi hitrega razvoja v industriji 4.0. Ker so senzorji postali manjši, cenejši ter strojno učinkovitejši, se vse pogosteje v industriji pojavlja zanimanje za prediktivno vzdrževanje. Cilj diplomske naloge je opisati ter predstaviti algoritme strojnega učenja, ki se lahko uporabijo pri izdelovanju tehnologij za prediktivno vzdrževanje. Predstavljeni sta tudi strojna ter programska oprema, ki sta že na voljo na trgu, ena izmed tehnologij pa je podrobneje opisana ter uporabljena na eksperimentu. V diplomskem delu je v uvodnih poglavjih predstavljen razvoj tehnik vzdrževanja in modeli napovedovanja pri prediktivnem vzdrževanju. Za tem sledi predstavitev strojnega učenja ter različnih kategorij le-tega. V tem poglavju so podrobneje opisani še najpogosteje uporabljeni algoritmi iz teh kategorij. Sledi pregled nekaterih rešitev strojne in programske opreme, ki je trenutno na voljo za uporabnike. Na koncu pa sem se odločil uporabiti eno izmed rešitev, in sicer iComox pametno napravo, ki sem jo uporabil na eksperimentu. Preizkus opreme je bil izveden v domačem okolju ter je tudi podrobno opisan, v zaključku pa je podano mnenje, ali je bila izbira te naprave smiselna za izbran eksperiment.

Keywords

prediktivno vzdrževanje;strojno učenje;pametni senzorji;umetna inteligenca;visokošolski strokovni študij;Aplikativna elektrotehnika;diplomske naloge;

Data

Language: Slovenian
Year of publishing:
Typology: 2.11 - Undergraduate Thesis
Organization: UL FE - Faculty of Electrical Engineering
Publisher: [L. Štajdohar]
UDC: 004.85:62-7(043.2)
COBISS: 129829379 Link will open in a new window
Views: 20
Downloads: 6
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: Internet of things for the predictive maintenance of the industry machines
Secondary abstract: Nowdays, we have more and more sensors and smart edge devices around the world, thanks to the rapid developments in the Industry 4.0. As sensors have become smaller, cheaper and more powerful, there is growing interest in predictive maintenance in industry. The aim of this diploma thesis is to describe nad present machine learning algorithms that can be used in the design of predictive maintenance technologies. Hardware and software products that are available on the market are presented and then one of the technologies is described in more detail and is then used in an experiment. In the introductory chapters of the thesis, the development of maintenance techniques and prediction models for predictive maintenance are presented. This is followed by an introduction to machine learning and its different categories. This chapter goes on to describe in more detail the most commonly used algorithms in these categories. It is then followed by an overview of some of the hardware and software solutions that are currently available to users. At the end, I decided to use one of the solutions, the iComox smart box and I used it in an experiment. The test of the equipment was carried out in a home environment and is described in detail. The conclusion concludes with an opinion on whether the choice of this device was reasonable for the chosen experiment.
Secondary keywords: predictive maintenance;machine learning;smart sensor;artificial intelligence;
Type (COBISS): Bachelor thesis/paper
Study programme: 1000315
Embargo end date (OpenAIRE): 1970-01-01
Thesis comment: Univ. v Ljubljani, Fak. za elektrotehniko
Pages: 37 str.
ID: 17111299