PhD position

10231


Integrated coupling of physical models and deep learning neural networks to predict the behavior of complex dynamic systems: application to refrigeration systems 

92160 

INRAE presentation

The French National Research Institute for Agriculture, Food, and Environment (INRAE) is a major player in research and innovation. It is a community of 12,000 people with 272 research, experimental research, and support units located in 18 regional centres throughout France. Internationally, INRAE is among the top research organisations in the agricultural and food sciences, plant and animal sciences, as well as in ecology and environmental science. It is the world’s leading research organisation specialising in agriculture, food and the environment. INRAE’s goal is to be a key player in the transitions necessary to address major global challenges. Faced with a growing world population, climate change, resource scarcity, and declining biodiversity, the Institute has a major role to play in building solutions and supporting the necessary acceleration of agricultural, food and environmental transitions.

Work environment, missions and activities

 Description of the subject

Today, the refrigeration sector uses around 20% of the world's total electricity consumption and is responsible for 8% of greenhouse gas emissions. In the current context of global warming, the need for refrigeration (food preservation, air conditioning, etc.) is expected to increase. The application of the Internet of Things (IoT) and artificial intelligence (AI) tools to the refrigeration industry is opening up great potential in terms of control and forecasting. However, applications of machine learning or AI methods are still rare in the refrigeration sector. The main difficulty lies in combining knowledge from two different fields: physical models in the field of refrigeration and AI models. This PhD thesis brings together skills in energy analysis of refrigeration systems (FRISE-INRAE), AI, Digital Twin (DT) and IoT (DVRC).

The main objective of this thesis is to develop a digital twin (DT) of a refrigeration enclosure. This DT will be built thanks to the coupling of knowledge in the two fields of AI and physics, in particular the integrated coupling of physical and deep learning models. It allows predicting the behaviour of the system (enclosure, products and refrigeration machine) during changes in operating conditions (outside temperature, product loading, breakdowns) using data collected by various sensors and knowledge from physical models to adapt the structure of the learning models themselves. Two coupling approaches will be studied in this thesis. First, we will look at "PINN" physics-informed neural networks, a class of machine learning methods that integrate physical knowledge into the neural network learning process. Another coupling method consists in using neural networks to adjust the parameters of physical models of refrigeration equipment, in particular kinetic and heat transfer models. The complexity of the physical equations governing this type of equipment makes this approach highly appropriate.

The DT developed in this work will also enable us to visualize future problems and the impact of a user decision on the system, and to propose solutions/scenarios to optimize its performance. This work will initially be developed for a specific refrigerated enclosure (cold room) made available to carry out the experiments.

The final objective is to have a 'generic' tool that could be adapted to various refrigeration systems.

Working environment

The Institut national de recherche pour l'agriculture, l'alimentation et l'environnement (INRAE) is a public research institute that brings together a working community of 10,000 people, with 273 research, service and experimentation units, spread across 18 centers throughout France. INRAE is a world leader in agricultural and food sciences, plant sciences and animal sciences. Its research aims to develop solutions for multi-performance agriculture, high-quality food and sustainable management of resources and ecosystems. FRISE-INRAE research unit, based in Antony (92), studies cooling and refrigeration systems for food preservation with low environmental impact and food safety. The unit comprises 3 teams, including 2 research teams, Enerfri (Refrigeration and Energy) and Metfri (Refrigeration and Food), and 1 technical support team, Techfri. The Enerfri research team (Hong-Minh Hoang, Anthony Delahaye) which will host the thesis works on reducing the energy consumption and the environmental impact of refrigeration systems.

The De Vinci Research Center (DVRC) based at La Défense brings together the teaching and research staff of the Pôle Léonard de Vinci schools: the Ecole de Management (EMLV) and the Ecole d'Ingénieurs (ESILV). Its areas of expertise contribute to the cluster's strategic positioning in terms of innovation, digitalization, transversality and ecological transition. DVRC is composed of 3 interdisciplinary axes:

- MISTIC "New materials, intelligent systems and innovative companies", and

- 2EMARK "Energy efficiency and socially responsible markets".

- STARCS "Data science, digital transformation, risks and complex systems" (Nedra Mellouli)

Interdisciplinarity and responsible research are closely linked in the organization of these areas. Indeed, integrating environmental or societal concerns into research work means mobilizing multiple skills and expertise to establish an interdisciplinary dialogue conducive to resolving complex issues

Thesis direction and funding

The thesis, funded by DVRC and Inrae, will be supervised by Nedra Mellouli, Hong-Minh Hoang and Anthony Delahaye within the Interfaces doctoral school (ED 573) at Paris-Saclay university

Training and skills

Master's degree/Engineering degree

  • Scientific: Master's degree in computer science, mathematics, process engineering, energy. Development skills in Python/PyTorch, R, Matlab
  • A liking for:
    • Machine learning (AI) methods, IoT, Digital Twin
    • Analysis of physical phenomena in energy systems
  • Skills:
    • Autonomy, knowledge of project management
    • Communication skills and knowledge transfer

INRAE's life quality

By joining our teams, you benefit from (depending on the type of contract and its duration):

- up to 30 days of annual leave + 15 days "Reduction of Working Time" (for a full time);
parenting support: CESU childcare, leisure services;
- skills development systems: trainingcareer advise;
social support: advice and listening, social assistance and loans;
holiday and leisure services: holiday vouchers, accommodation at preferential rates;
sports and cultural activities;
- collective catering.


OFFER REFERENCE

  • Contract: PhD position
  • Duration: 36 months
  • Beginning: 01/12/2024
  • Remuneration: 2100€ gross monthly
  • Reference: OT-22283
  • Deadline: 01/09/2024





INRAE - National Research Institute for Agriculture, Food and the Environment



Visit employer page


Deadline: 2024-09-01
Location: France, Antony
Categories: Computer Engineering, energy engineering, Machine Learning, Mathematics, PhD, Physics, Process Engineering,

Apply


Ads
Technische Universität Graz


GFZ - Helmholtz / Deutsches GeoForschungsZentrum


AIT Austrian Institute of Technology


Max Planck Institute for Social Law and Social Policy


Helmut-Schmidt-Universität - Universität der Bundeswehr Hamburg


BOKU Universität für Bodenkultur Wien


SAL Silicon Austria Labs GmbH


Helmholtz Centre for Infection Research (HZI)


More jobs from this employer