Betreuer: | Ban-Sok Shin, Shengdi Wang |
Art der Arbeit: | Masterarbeit (MSc) |
Arbeit beendet: | 07/2021 |
Bearbeiter: | Zhengyu Chen |
Status: | abgeschlossen |
ANT-Signatur: | |
Kurzfassung: | Your mission: Seismic exploration is of great importance for future planetary missions: the planet’s subsurface structure can reveal a great deal of information about the planet’s past and present. We envision the use of a network of multiple agents – of a swarm – to autonomously and cooperatively reconstruct the subsurface structure. To compute an image of the subsurface typically several processing steps of seismic measurements are required. One of those steps is a so called normal moveout strategy that can be used to obtain velocity estimates of the subsurface. At present seismic exploration surveys such data processing is conducted in a centralized manner, where all seismic measurements are available at a single computing entity. However, for application in a network of multiple agents these computations need to be conducted in a distributed fashion. In this thesis, the task of the potential candidate is to implement the normal moveout strategy in a distributed fashion for an estimation of the subsurface velocities and layer depths within a network of geophones. To this end, layered Earth models shall be simulated by solving a wave equation that generates synthetic measurements at the geophone locations. Based on the seismic measurements the objective is to reconstruct the velocities and depths of the layered Earth model in a distributed fashion. The developed scheme shall be tested via simulations with both synthetic and real measurements. If time allows, experiments in a real environment shall be conducted to validate the simulation results. Your qualifications: - Solid background in signal processing - Good programming skills in Python or Matlab - Independent and self-motivated working style - Motivativation to solve complex tasks |