Vous êtes ici : Accueil » Kiosque » Annonce

Identification

Identifiant: 
Mot de passe : 

Mot de passe oublié ?
Détails d'identification oubliés ?

Annonce

5 janvier 2017

Self-developing Fault Diagnostic and Prognostic Tools for the Prescriptive Maintenance of Offshore Wind Turbines


Catégorie : Stagiaire


Context and aims

This Master internship subject is part of a collaboration between the High National Engineering School of Mines (Douai/France) and the French Atomic Energy Commission (CEA) of Saclay (Paris/France). The first laboratory develops fault diagnostic and prognostic tools based on the use of machine learning and data mining techniques while the second laboratory offers its ExpressIFTM software for the capitalization of human expertise by fuzzy inference system. This capitalization can be a direct expression of knowledge or derived from learning algorithms capable of retrieving the knowledge of the expert transcribed in the form of "If-Then" rules in datasets. These rules would help human operators of supervision to build online reasoning, verify hypotheses, etc. in order to quickly converge towards the most appropriate and effective decision-making according to the situation’s conditions.

An online adaptive machine learning and data mining scheme in order to perform an early fault diagnosis of critical components in wind turbines was developed in the High National Engineering School of Mines of Douai-France. The objective of this Master internship is to integrate the ExpressIFTM tool into the developed scheme in order to provide explaining decision under comprehensive form (natural language). This explanatory capacity would improve the situation awareness and understanding by human operators and thus assist them in their decision-making to plan the appropriate maintenance operations at the lower cost. To do this, it will be necessary to adapt the outputs of the diagnostic module so that it can be used in ExpressIFTM to produce a set of rules in natural language comprehensible by the human operators of supervision.

A PhD thesis will be proposed at the end of this Master internship in order to solve the challenges related to the size of the system (large wind farm), the large volume of data, their speed and their integrity (Big data challenges).

Internship Location

The work of this internship will be done at the Computer Science and Automatic Control Department of the High National Engineering School of Mines in Douai-France under the supervision of Prof. Moamar Sayed Mouchaweh.

Candidate profile

Master degree or equivalent in Computer Science or equivalent domain. The candidate must demonstrate a scientific expertise in one or more areas of: Machine learning and data mining, data processing, fault diagnostics, fault prognostics, modeling of complex dynamic systems. The candidate must have solid experience in programming using C, Python or Matlab.

Contact persons

Applicants should submit a curriculum vitae with the names and email addresses of two references to:

Dans cette rubrique

(c) GdR 720 ISIS - CNRS - 2011-2015.