Vibrations here to stresses there: virtual sensing methods for offshore wind turbines monitoring

IFP Energies nouvelles - Direction Physico-chimie et Mécanique appliquées

Stage Data / Mathématiques Appliquées Hauts-de-Seine entre mars et avril 2023 6 mois


IFP Energies nouvelles (IFPEN) est un acteur majeur de la recherche et de la formation dans les domaines de l’énergie, du transport et de l’environnement. De la recherche à l’industrie, l’innovation technologique est au cœur de son action, articulée autour de quatre priorités stratégiques : Mobilité Durable, Energies Nouvelles, Climat / Environnement / Economie circulaire et Hydrocarbures Responsables.

Dans le cadre de la mission d’intérêt général confiée par les pouvoirs publics, IFPEN concentre ses efforts sur :

  • l’apport de solutions aux défis sociétaux de l’énergie et du climat, en favorisant la transition vers une mobilité durable et l’émergence d’un mix énergétique plus diversifié ;
  • la création de richesse et d’emplois, en soutenant l’activité économique française et européenne et la compétitivité des filières industrielles associées.

Partie intégrante d’IFPEN, l’école d’ingénieurs IFP School prépare les générations futures à relever ces défis.

Vibrations here to stresses there: virtual sensing methods for offshore wind turbines monitoring

2023 Master's degree research internship (6 months) - IFPEN

One of the ways to reduce the cost of wind energy, especially in an offshore context, is to optimize the maintenance of wind farms. It is therefore desirable to be able to track the health of wind turbines remotely, using the monitoring data available (typically vibration data from accelerometers), to anticipate maintenance and failures.

This monitoring can be purely data-based, using machine learning approaches that try to quantify structural damages directly from sensor time series. These approaches are promising but lack physical grounds. There also exists simulators that can calculate, on a sound physical basis, many quantities that are difficult to access with sensors only (mechanical stresses at critical points of offshore wind turbine structures for instance). However, simulators only offer an approximate description of reality. It is therefore natural to want to bring together simulators and data and to benefit from the best of both worlds: data allows to correct the simulator, while the simulator allows to extrapolate data.

Within this paradigm, the objective of the internship is to explore the concept of virtual sensing: from time series measured from installed sensors on the structure, how to estimate related time series at another location of the structure by means of a model (simulator) of the asset?

For instance, from acceleration measurements recorded in an offshore wind turbine tower, we aim to predict mechanical stress time series in the foundation of the jacket structure that supports the turbine, for they are critical for the structural integrity (fatigue lifetime) of the asset.

Such data extrapolation can be done by using various methods that relate different locations and/or physical quantities of the structure. They can be based on purely mechanical principles (exploitation of the modal basis of the structure); statistical methods (neural networks “constrained” by the simulator) [2] or signal processing (Kalman filter acting as a data assimilation tool) [3].

The intern will have access to an aero-hydro-servo-elastic model of an offshore wind turbine, which he/she will use to generate synthetic data and test virtual sensing methods among those mentioned above. It will be assumed in this work that the simulator has already been calibrated with the real asset it models. The quality of the estimated data will be assessed by comparing the predictions with reference data (generated by the simulator), allowing to test different settings (e.g. position and number of sensors, operating conditions, position and type of quantities of interest).

Tasks & objectives

  • Take control of the aero-hydro-servo-elastic model of an offshore wind turbine to generate data;
  • Literature review of the available methods and implementation of selected techniques;
  • Comparison of the different virtual sensing techniques for different settings, sensitivity/parametric studies.

Supervision context

This research internship is part of a collaborative project between IFPEN and Inria (French Institut national de recherche en sciences et technologies du numérique) on wind turbine monitoring. The intern will therefore have the opportunity to interact with engineers from both institutes.

Profile and skills

M2 student or last year of engineering school in mechanical engineering (knowledge in structural vibrations) and/or applied mathematics (knowledge in statistics/data science).

  • Experience with Python programming and basic knowledge of the wind energy field would be an asset.

Duration of the internship : 6 months, starting end of March 2023
Location : IFPEN, Rueil-Malmaison (92), France
Advisory team : JL. PFISTER, M.R. EL AMRI, L. MEVEL (Inria), E. DENIMAL (Inria)

References
[1] Augustyn et al. “Feasibility of modal expansion for virtual sensing in offshore wind jacket substructures,” Marine Structures 79 (2021)
[2] Li & Zhang, “Physics-informed deep learning model in wind turbine response,” Renewable Energy 185 (2022)
[3] Maes al. “Dynamic strain estimation for fatigue assessment of an offshore monopile wind turbine using filtering and modal expansion algorithms,” Mechanical Systems and Signal Processing 76-77 (2016)


(PDF - Max : 5 Mo)
(PDF - Max : 5 Mo)

En cliquant sur postuler, je crée mon profil iQuesta et j'accepte les conditions d'utilisation d'iQuesta.
contact

IFP Energies nouvelles - Direction Physico-chimie et Mécanique appliquées
Jean-Lou Pfister

Indemnité Oui

9 Annonces
Métiers
Data Analyst
Le Data Analyst occupe une place stratégique dans un monde de plus en plus numérique où la quantité de données…
Conseils
Pour que la vie d’étudiant ne rime pas avec parcours du combattant, retrouvez tous nos conseils stage / alternance / emploi