de | en

Thomas Pöschl

Lehrveranstaltungen und Termine

Ausgeschriebene Angebote für Abschlussarbeiten

Development of a Neural Network for Online Event Reconstruction for a Radiation Monitor

The Multi-purpose Active-target Particle Telescope (MAPT) is a newly developed radiation detector for space applications. The detector shall be used to monitor the radiation environment on spacecraft and satellites. It is most sensitive to low-energy protons and ions and can distinguish the particles by their interactions with the material of the detector.

In this thesis, a neural network shall be developed that can reconstruct the direction, energy, and species of the measured particle in near real time. You will have to design a network structure, train the network with simulation data, and assess its performance. The trained network shall then be ported to on a small on-board computer that can be implemented in MAPT.     

Tasks

  • Acquire necessary theoretical understanding of neural networks and the different architectures.
  • Implement a neural network in Python using existing libraries.
  • Train and validate the algorithm with detector data generated with the high-energy physics simulation tool Geant4.
  • Compare the performance of the network to existing analysis approaches.

Prerequisites

Experience in Python and C++ programming is helpful, but not required. An introductory course on C++ programming is offered.

 

Contact

Thomas Pöschl, Room PH1 3554, Thomas.poeschl@ph.tum.de

Prof. Stephan Paul, Room PH1 3263, stephan.paul@tum.de

geeignet als
  • Masterarbeit Kern-, Teilchen- und Astrophysik
Themensteller(in): Stephan Paul
Nach oben