diplomsko delo Visokošolskega strokovnega študijskega programa I. stopnje Strojništvo
Abstract
Osrednji problem in področje raziskovanja tega dela je strojno učenje kot pripomoček za ugotavljanje napak na strojnih elementih. V prvem delu je obravnavana raziskava, iz katere so pridobljeni surovi podatki in njihova pred-obdelava v ustrezno obliko. V omenjeni raziskavi se opazuje 5 različnih napak na ležajih: aksialna in radialna preobremenitev, preobremenitev upogibnega momenta, kontaminacija in napaka kletke. V naslednjih sklopih so predstavljene teoretične osnove strojnega učenja, algoritmi za uspešno analizo ter primeri uporabe na konkretnih podatkih. Kot pomemben del raziščemo tudi optimizacijo parametrov pri različnih modelih in obravnavamo korektnost dobljenih rezultatov
Keywords
diplomske naloge;strojno učenje;umetna inteligenca;kakovost;ležaj;Python;scikit-learn;sklearn;
Data
Language: |
Slovenian |
Year of publishing: |
2018 |
Typology: |
2.11 - Undergraduate Thesis |
Organization: |
UL FS - Faculty of Mechanical Engineering |
Publisher: |
[K. Kubelj] |
UDC: |
004.85:621.82(043.2) |
COBISS: |
16251931
|
Views: |
988 |
Downloads: |
454 |
Average score: |
0 (0 votes) |
Metadata: |
|
Other data
Secondary language: |
English |
Secondary title: |
identification of bearing faults with machine learning packet scikit-learn |
Secondary abstract: |
The main focus and research field of this work is machine learning as a tool for classifying faults of machine elements. In the first part, we address the research, from which we take the raw data and the preprocessing of the gathered data set. The research takes a look at 5 different bearing faults: axial and radial overload, bending moment, contamination and shield defect. Next, we take a look at the theoretical background of machine learning, algorithms for analysis and examples of practical use. As an important aspect we research the possibilities of optimizing model parameters and evaluate the success of our predictions. |
Secondary keywords: |
machine learning;artificial intelligence;quality;bearing;Python;scikit-learn;sklearn; |
Type (COBISS): |
Bachelor thesis/paper |
Study programme: |
0 |
Thesis comment: |
Univ. v Ljubljani, Fak. za strojništvo |
Pages: |
XXII, 41 str. |
ID: |
10956367 |