Abstract

On the basis of a collaboration between the French Ministry of Higher Education, Research and Innovation (MESRI), the German Federal Ministry of Education and Research (BMBF), Fraunhofer ISE, CEA, LPNC, Stiebel-Eltron GmbH & Co. KG, EDF R&D – Dept. Technology and Research for Energy Efficiency.

Coordinators of the project

• Martial Mermillod, LPNC & Head of the MIAI Chaire Towards Robust and Understandable Neuromorphic Systems. Laboratoire de Psychologie & NeuroCognition
• Lilli Frison, Constanze Bongs - Fraunhofer Institute for Solar Energy Systems
• Marina Reyboz, Cédric Gouy-Pailler – CEA Grenoble
• Johannes Brugmann & Michael Schaumlöffel Stiebel Eltron - GmbH & Co. KG
• Thuy-an Nguyen - EDF R&D Dept. Technology and Research for Energy Efficiency

Description

This project will address the development of novel machine learning methods, which are necessary in order to make sure that Artificial Neural Networks learn autonomously and adaptively in a changing environment without having to store the complete past data nor forgetting past knowledge. This requires research in the field of incremental and online learning for time series data (CEA, LPNC). Consequently, the developed novel AI methods will be applied to adaptive HP control and supervision (ISE) and brought into practice, starting from virtual testing, followed by laboratory testing in a HIL-environment (EDF) up to their application in an industry-standard heat pump controller (Stiebel Eltron). The latter will be validated in a real-life pilot installation.

Budget of the project

~ 1586 k€ for 3 years (2021-2024).