Institut Français des Sciences et Technologies des Transports, de l'Aménagement et des Réseaux

Type de document

THESE

Langue

anglais

Auteur

CHAMROUKHI

Résumé / Abstract

This research addresses the problem of diagnosis and monitoring for predictive maintenance of the railway infrastructure. In particular, the switch mechanism is a vital organ because its operating state directly impacts the overall safety of the railway system and its proper functioning is required for the full availability of the transportation system; monitoring it is a key task within maintenance team actions. To monitor and diagnose the switch mechanism, the main available data are curves of electric power acquired during several switch operations. This study therefore focuses on modeling curve-valued or functional data presenting regime changes. In this thesis we propose new probabilistic generative machine learning methodologies for curve modeling, classification, clustering and tracking. First, the models we propose for a single curve or independent sets of curves are based on specific regression models incorporating a flexible hidden process. They are able to capture non-stationary (dynamic) behavior within the curves and address the problem of missing information regarding the underlying regimes, and the problem of complex shaped classes. We then propose dynamic models for learning from curve sequences to make decision and prediction over time. The developed approaches rely on autoregressive dynamic models governed by hidden processes. The learning of the models is performed in both a batch mode (in which the curves are stored in advance) and an online mode as the learning proceeds (in which the curves are analyzed one at a time). The obtained results on both simulated curves and the real-world switch operation curves demonstrate the practical use of the ideas introduced in this thesis.

Editeur

Université de technologie de Compiègne