District heating networks (DHNs) are critical for integrating renewable and waste heat, supporting efforts to combat global warming. In France, the low-carbon strategy focuses on expanding DHNs, including smaller networks with multiple heat sources like solar thermal, and storage. Advanced control methods, such as model-predictive control (MPC), aim to replace manual practices but face challenges due to high deployment costs and technical complexity, particularly for modest-sized networks.

Large DHNs use optimization tools like MILP for real-time production planning, while smaller ones rely on simpler, rule-based systems. Research explores reducing the complexity of MPC through offline modeling, systematic methods, and machine learning, comparing strategies for adaptability, performance, and operational efficiency.

This postdoctoral project seeks to develop, test, and deploy innovative control strategies for DHNs using physical simulators and real-world implementation. Objectives include cost optimization, robustness, and ease of deployment, with results disseminated through publications, conferences, and potential patents. The researcher will access advanced tools, computational resources, and experimental facilities (a real scale multi-source district heating network at CEA) to achieve these goals.

What you will do:

The detailed objectives for this 24-months postdoctoral project are:

· Establish a state-of-the-art of advanced strategies for production control of DHNs, including (but not limited to) operations research or machine learning-inspired techniques

· Specify, implement and test off-line the most promising approaches using a physical simulator, and considering difference performance indicators (operation costs reduction, robustness, with respect to imperfect previsions).

· Deploy the most promising control strategies online on the real experimental DHN of CEA-INES (specification and implementation of the control chain on the automation hardware, continuous performance assessment/comparison)

· Disseminate the result through scientific publications and participations to international conferences, possibly after patenting.

What you bring to the table:

· PhD in Optimization and/or Simulation

· Strong interest / background for the energy sector

Knowledge of C++/Python programming
Knowledge/Interest in Machine/Deep Learning

You bring creativity, enjoy teamwork and can undertake research projects autonomously.

What you can expect:

In the heart of the Alps, an exceptional and internationally recognised scientific ecosystem, in a city committed to the environmental transition "Grenoble European Green Capital 2022",

A balance between private and professional life (50 days off/year), an active Works Council for leisure and extra-professional activities.

Salary according to your qualifications and experience, access to a savings scheme funded by the CEA,

Industrial-grade facilities for research

Participation to scientific conferences.

Training to strengthen your skills or acquire new ones