Master Thesis: Neural networks heat transfer prediction for two-phase closed thermosyphons

25. Mai 2022

Master Thesis

Objective: 

The task of the work includes the training of an artificial neural network (ANN) based on experimental data for the heat transfer prediction at two-phase closed thermosyphons (TPCT). 

Background:

TPCTs are highly efficient passive heat transfer devices that can continuously transport heat from a heat source to a heat sink without any mechanical and electrical components. The thermal performance of TPCTs depends on the thermo-fluid dynamic behaviour in the evaporator and condenser sections. However, the prediction of the heat transfer in these sections might represent an issue to be addressed, since the available correlations (Fig. 1) in the literature are valid only for the range of parameters covered in the respective study. The goal of this work is to train an ANN (Fig. 2) using data points of experimental results to predict the heat transfer of TPCTs. 

Procedure: 
  • Familiarization with the fundamentals of TPCTs and ANN,
  • Literature review on experimental works with TPCTs,
  • Training of the ANN,
  • Validation with separated data from training,
  • Consideration of counter overfitting and model generalization,
  • Evaluation and classification of the results,
  • Written elaboration and oral presentation.
Requirements: 
  • Interest in deep learning,
  • Knowledge in math and stats,
  • Basic knowledge/experience with Matlab or Python (e.g. TensorFlow),
  • Very good German or English skills.
Kontakt

M.Sc. Sergio Cáceres / M.Sc. Marc Kirsch
Pfaffenwaldring 31 • Room no. 2.232, D-70569 Stuttgart


sergio.caceres@ike.uni-stuttgart.de
+49 711 685 69662

mark.krisch@ike.uni-stuttgart.de
+49 711 685 61798

Datei-Anhänge

Kontakt

Dieses Bild zeigt Sergio Iván Cáceres Castro

Sergio Iván Cáceres Castro

M.Sc.

Wissenschaftlicher Mitarbeiter / Doktorand

Dieses Bild zeigt Marc Kirsch

Marc Kirsch

M.Sc.

Wissenschaftlicher Mitarbeiter / Doktorand

Zum Seitenanfang