DEEP

Data-driven Embedded Echosystem for Propulsion and Power applications

---

Applications

• Data-driven model for single-stage centrifugal compressor
• Data-driven model to replace expensive non-linear equation of state evaluations in CFD computations

---

Developers

• A. Giuffre', PhD Researcher, Propulsion & Power, TU Delft
• E. Bunschoten, PhD Researcher, Propulsion & Power, TU Delft
• A. Cappiello, Post-Doc Researcher, Propulsion & Power, TU Delft
• M. Majer, PhD Researcher, Propulsion & Power, TU Delft
• M. Pini, Assistant Professor, Propulsion & Power, TU Delft

---

Pre-requisites

Python 3.x

NumPy

SciPy

TensorFlow

PyTorch

GPyTorch

scikit-learn

MatPlotLib

Installation with conda is recommended