Digital Twin and Explainable AI for Energy Optimization in Agri-PV Viticulture

Digital Twin and Explainable AI for Energy Optimization in Agri-PV Viticulture

Réf ABG-134961

Stage master 2 / Ingénieur

Durée 36 mois

Salaire net mensuel ~2 300 € brut/mois

12/01/2026

Université de Haute alsace

Lieu de travail

MULHOUSE Grand Est France

Champs scientifiques

• Indifférent

Mots clés

artificial intelligence, big data, simulations, digital twins

Date limite de candidature

12/02/2026

Établissement recruteur
Site web :

Project partners:

• IRIMAS Laboratory, University of Haute Alsace
• EDF (Electricité de France)
• Wolfberger (Alsatian wine cooperative)
• European program INTERREG Agri-PV
• German and Swiss partners of the project

Description

Context

This PhD thesis is part of the European Agri-PV project (Interreg Upper Rhine, 2025–2028): https://www.interreg-rhin- , which explores agrivoltaics in viticulture as a solution to climate and energy challenges. This cross-border initiative brings together French, German, and Swiss academic and industrial partners who test different photovoltaic systems and study their impacts on viticulture, electricity yield, and landscape.

Three pilot sites (in Rhineland-Palatinate, Baden-Württemberg, and Switzerland) are equipped with Agri-PV systems and gather data (PV production, microclimate, vine development, etc.) to design monitoring tools and practical guides for winegrowers and local authorities.

By promoting dual land use, Agri-PV aims to diversify winegrowers' income, enhance the resilience of local electricity grids, and support the regional energy transition. The key challenge is to leverage these interdisciplinary datasets and systems (PV, agriculture, energy grids) to optimize self-consumption and renewable energy management in Upper Rhine vineyards.

Thesis Objectives

• Multi-physics Digital Twin: Design a high-fidelity virtual model of vineyard-based Agri-PV installations, capturing the behavior of the PV/electric system. The digital twin will simulate various control scenarios and optimize PV configuration and operations.
• Explainable Artificial Intelligence (XAI): Develop and adapt explainable machine learning algorithms to analyze the experimental datasets (weather, PV, soil, vine, etc.). XAI techniques will enhance model transparency and user confidence, especially for winegrower-facing applications.
• Real-Time Modeling and Simulation: Build an integrated energy model (PV production, storage, local consumption) for the pilot sites and validate it using the OPAL-RT real-time simulator at IRIMAS Institute. Predictive simulations will support optimal strategy design (e.g., storage charge/discharge, inverter control).
• Energy Coupling Analysis: Investigate the interaction between PV production, energy storage, and local consumption across transnational vineyard sites. The goal is to propose robust and reproducible energy management scenarios tailored to viticulture.

PhD Missions

• Participate in measurement campaigns on the pilot Agri-PV sites (France, Germany, Switzerland).
• Handle preprocessing, annotation, and structuring of the transboundary experimental database in collaboration with research partners and wine institutes.
• Develop the digital twin jointly with technical partners and integrate it into simulation platforms (e.g., Matlab/Simulink).
• Implement XAI algorithms on the collected datasets to extract interpretable rules (feature extraction, impact visualization), ensuring reproducibility and robustness of the machine learning models.
• Validate models and control strategies using laboratory infrastructure: test scenarios on the OPAL-RT real-time simulator, and conduct physical tests on IRIMAS' smart microgrid platform.
• Collaborate with a multidisciplinary team (agronomists, energy experts, climatologists) and contribute to the scientific output of the project (publications, reports, Agri-PV practical guides).

Profil

• Knowledge of system modeling and simulation.
• Proficiency in AI and machine learning techniques.
• Strong analytical skills, scientific rigor, and autonomy. Enjoyment of field work and teamwork in interdisciplinary settings.

Prise de fonction

02/03/2026

---