Vers une approche générique de prévision de la production renouvelable et de la demande à l'échelle locale en utilisant des sources de données multiples
Vers une approche générique de prévision de la production renouvelable et de la demande à l'échelle locale en utilisant des sources de données multiples
Seamless forecasting of local energy production and demand using multiple heterogeneous data sources
Spécialité
Energétique et génie des procédés
Ecole doctorale
ISMME - Ingénierie des Systèmes, Matériaux, Mécanique, Énergétique
Directeur de thèse
KARINIOTAKIS Georges
Co-directeur
CAMAL Simon
Unité de recherche
Energétique et Procédés
Contact
Date de validité
30/06/2024
Date de début de thèse
01/08/2024
Site Web
Mots-clés
Digitalisation de l'énergie, data science, smart grids, prévision, optimisation, préservation de la confidentialité
Energy digitalisation, data science, smart grids, forecasting, optimisation, privacy-preserving
Résumé
Context and challenges:
Short-term energy forecasting for the next minutes to days ahead, is a prerequisite for the economic and safe operation of modern power systems and electricity markets especially under high renewable energy sources (RES) penetration. The different contexts of application make that end-users require models that have a broad number of properties especially when they are applied operationally. They should cover multiple time frames (from minutes to days ahead) and multiple RES technologies (i.e. wind, solar, hydro) as well as their aggregations (i.e. in the form of virtual power plants – VPP). They should use as input the very large amount of data available, while dealing efficiently with dimensionality. The data sources may be measurements from the power plants, various types of satellite images, sky camera images, various feeds of numerical weather prediction and others. They should be generic enough to be easily replicable to different sites or demand forecasting. They should also be resilient against imperfect or corrupted data streams; be interpretable enough; and be able to deal with structural changes in the physical system (e.g. addition of assets to a VPP or equipment in a smart home). So far separate models are developed for each of these aspects. The thesis is realized in the frame of the PEPR TASE project Fine4Cast coordinated by the supervisors of this thesis. PERSEE has an international visibility in the field of energy forecasting thanks to a long track of national and European projects, PhDs and publications in the area.
Main objective of the thesis:
This thesis will develop a seamless forecasting approach for net-load and joint load and renewable production that meets the above requirements, while being at least as accurate as the currently used partial models. It will also preserve privacy of the different data sources. The modelling approach should be probabilistic giving the possibility to estimate the uncertainty in the forecasts. Combination methods of probabilistic forecasts will be assessed.
Methodology and expected results:
A seamless method has been proposed by PERSEE that optimally combines the available data sources to derive a probabilistic forecast of RES production at multiple temporal scales and aggregation levels. Adapting this seamless concept to local demand or net-load has not yet been proposed in the literature. The methodology will start by identification of adequate explanatory variables from multiple data sources (multiple weather predictions and simulations, local measurements, multiple types of satellite-based images, etc.). The second step will ensure the scalability of the forecasting approach to large dimensions and the adaptivity to structural change in the production and demand assets. Validation will be done using available real-world data sets. Emphasis will be given on assessing the contribution of each available data source in a cost-benefit analysis context.
Contexte
...
Encadrement
Quotités d'encadrement:
Georges Kariniotakis 40%
Simon Camal 30%
Profil candidat
PROFILE:
Engineer and / or Master of Science degree (candidates may apply prior to obtaining their master's degree. The PhD will start though after the degree is succesfully obtained).
Good level of general and scientific culture. Good analytical, synthesis, innovation and communication skills. Qualities of adaptability and creativity. Motivation for research activity. Coherent professional project. Skills in programming (eg R, Python, Julia,…). A succesful candidate will have a solid background in three or more of the following competencies:
• applied mathematics, statistics and probabilities
• data science, machine learning, artificial intelligence
• energy forecasting
• power system management, integration of renewables
• optimisation
Expected level in french : bon niveau souhaitable
Expected level in english : excellent
Please send the following elements by email (in pdf format) to Prof. George Kariniotakis (georges.kariniotakis@minesparis.psl.eu) AND to Dr. Simon Camal (simon.camal@minesparis.psl.eu):
• Curriculum vitae (CV).
• Motivational letter for the application (cover letter).
• Contact details of two individuals that can provide a letter of reference (and eventually available already letters of reference).
• Copy of grade transcripts and last diploma (in English or French).
Please use in the title of email the acronym of this PhD topic “PHD-2024-ERSEI-PreviSeamlessâ€
Deadline for applications: 02/29/2024. The position will remain open until a suitable candidate is found.
Do not hesitate to email to the above addresses for an early expression of interest and for further information on the position.
Références
...
Type financement
Financement d'un établissement public Français
Partenariat/contrat
Projet Fine4CAST ('Next Generation Energy Demand and Renewable Production Forecasting Tools for Fine Geographical and Temporal Scales').