polito.it
Politecnico di Torino (logo)

Rolling bearing damage characterization - A machine learning approach

Milad Rahmani Tootkaboni

Rolling bearing damage characterization - A machine learning approach.

Rel. Alessandro Fasana. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Meccanica (Mechanical Engineering), 2021

[img]
Preview
PDF (Tesi_di_laurea) - Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives.

Download (22MB) | Preview
[img] Archive (ZIP) (Documenti_allegati) - Altro
Licenza: Creative Commons Attribution Non-commercial No Derivatives.

Download (7MB)
Abstract:

Rolling bearing damage characterization - A machine learning approach This thesis is done based on the article published on Mechanical Systems and Signal Processing journal . Article name: The Politecnico di Torino rolling bearing test rig: Description and analysis of open access data. Article authors: Alessandro Paolo Daga, Alessandro Fasana, Stefano Marchesiello, Luigi Garibaldi . The research has been taken place and the test has been conducted in the Dynamic and Identification Research Group (DIRG) of the Department of Mechanical and Aerospace Engineering at Politecnico di Torino. The main article by PoliTo works on two main different tests, Variable Speed and Load test and Endurance test. However, on this thesis only the data of Variable Speed and Load test is used. The complementary information is available on: ftp://ftp.polito.it/people/DIRG_BearingData/ This thesis aims to clarify the defect classification of roller bearings. The first chapter gives an introduction on bearings. The main test done at Politecnico di Torino is briefly explained on the second chapter. The basic matrices formation which are used for further final models matrices is presented in chapter three. On chapter four different models and their different factors are completely defined. Data preparation for machine learning process such as train and test data sets are fully described in chapter five. Machine learning process and all the pertaining results are presented in chapter six. The concept of feature selection is fully described in chapter seven. In chapter eight the conclusion is presented. At the end the extra information is presented in four separated appendices.

Relatori: Alessandro Fasana
Anno accademico: 2020/21
Tipo di pubblicazione: Elettronica
Numero di pagine: 126
Soggetti:
Corso di laurea: Corso di laurea magistrale in Ingegneria Meccanica (Mechanical Engineering)
Classe di laurea: Nuovo ordinamento > Laurea magistrale > LM-33 - INGEGNERIA MECCANICA
Aziende collaboratrici: University of Massachusetts Dartmouth
URI: http://webthesis.biblio.polito.it/id/eprint/19539
Modifica (riservato agli operatori) Modifica (riservato agli operatori)