ATTENTION/WARNING - NE PAS DÉPOSER ICI/DO NOT SUBMIT HERE

Ceci est la version de TEST de DIAL.mem. Veuillez ne pas soumettre votre mémoire sur ce site mais bien à l'URL suivante: 'https://thesis.dial.uclouvain.be'.
This is the TEST version of DIAL.mem. Please use the following URL to submit your master thesis: 'https://thesis.dial.uclouvain.be'.
 

Exploring anomaly detection on time series for predictive maintenance

Vandermosten, Hugo
(2024)

Files

Vandermosten_25741900_2024.pdf
  • Closed access
  • Adobe PDF
  • 2.79 MB

Details

Supervisors
Delvenne, Jean-Charles
Degree label
Master [120] : ingénieur civil en science des données, à finalité spécialisée
Abstract
In this thesis, the subject of predictive maintenance is addressed. It is executed with a data-driven approach for anomaly detection. Therefore, multiple models and procedures to tackle this challenge such as ARIMA and vector-based models are presented. A set of processing methods for time series is also examined. The study was performed in collaboration with AGC Glass Europe, who provided continuous support and access to some of their data. The goal is to give a general overview of important aspects to take into account when exploring the data-driven approach. Each presented architecture is also tested and discussed with data from AGC to assert the relevance of explored models. Each analysis is also performed with a highly critical approach.