This thesis aims to build a flexible multi‑objective optimization framework for satellite constellation design, capable of trading coverage, revisit time, number of satellites, and cost. You will implement and benchmark evolutionary algorithms (e.g. NSGA‑II/III) and/or other multi‑objective methods, and connect them to simplified or high‑fidelity constellation performance models. The work extends classic constellation optimization approaches with modern software engineering and, if desired, digital‑twin ideas similar to those used in recent ML‑enabled frameworks.​

You will be free to choose your algorithmic and software stack (Python, Julia, C++, existing MOEA libraries, etc.), but you must justify these choices and design for extensibility (e.g. plug‑in objectives and constraints).

Mini demo challenge (attach repo link in your application):
Create a very small multi‑objective optimization example that:

• Uses a 2‑objective toy problem (e.g. simple analytic “coverage” vs. “cost” or a standard test function such as ZDT1).

• Generates ~50–100 candidate solutions and approximates a Pareto front (via your own GA or an optimization library).

• Produces a scatter plot and a data file (CSV/JSON) of the non‑dominated solutions.

• In the README, explain in a few sentences:

  • Why you chose a particular optimization library or implemented your own,

  • How you would extend this setup to real constellation design metrics and constraints.

The demo is intentionally small; the key signal is how you select, combine, and justify tools.

Suggested reading:

• Chao, S.-L. et al., “Satellite constellation design optimization via multiple-objective evolutionary computation”, J. Guidance, Control, and Dynamics.[pure.psu]

• Liu, J. et al., “Satellite Constellation Orbit Design Optimization with Combined Genetic Algorithm and Semianalytical Approach”, Advances in Astronomy.[onlinelibrary.wiley]

Application:
Interested candidates should submit their application to julius.pinsker@faps.fau.de including:

• Current transcript of records (grades).

• A concise CV (1 page).

• A link to the mini demo repository described above.

I look forward to receiving your innovative applications.

Kategorien:

Kontakt: