Skip to Main content Skip to Navigation
Theses

Création d'un jumeau numérique d'un système de réfrigération et validation expérimentale : application à la détection de fuites

Abstract : The refrigeration sector consumes about 17% of the total electricity used in the world. This percentage is not negligible and requires making cooling systems more energy efficient, and reducing their impact on the greenhouse effect. In particular, refrigerant leaks are a problem that must be solved. Indeed, they have a direct and indirect impact on the environment. Also, leaks impose additional repair costs. For these reasons, it is very important to employ reliable and fast leak detection methods that ensure real-time detection. In this thesis, we have proposed to develop a physical model that represents the cooling system. This model is validated and calibrated with experimental tests. Once the model is validated, it is used as a data generator. These data will be used in addition to the data acquired in real time to ensure the training of a statistical model on the normal operation of the machine. This learning model will then be used to detect leaks in case of deviation from normal operation defined from the measurement of several parameters sensitive to the variation of the refrigerant charge. As the physical model of the machine is used as a complement to the real machine and not to replace it, we will call it a "digital twin".
Document type :
Theses
Complete list of metadata

https://pastel.archives-ouvertes.fr/tel-03728391
Contributor : ABES STAR :  Contact
Submitted on : Wednesday, July 20, 2022 - 12:18:16 PM
Last modification on : Tuesday, August 2, 2022 - 3:09:03 AM

File

2021UPSLM067_archivage.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-03728391, version 1

Citation

Etienne Haddad. Création d'un jumeau numérique d'un système de réfrigération et validation expérimentale : application à la détection de fuites. Energie électrique. Université Paris sciences et lettres, 2021. Français. ⟨NNT : 2021UPSLM067⟩. ⟨tel-03728391⟩

Share

Metrics

Record views

36

Files downloads

10