| |
 |
| Paper: |
Using Heuristic Algorithms to Optimize Observing Target Sequences |
| Volume: |
485, Astronomical Data Analysis Software and Systems XXIII |
| Page: |
73 |
| Authors: |
Sosnowska, D.; Ouadahi, A.; Buchschacher, N.; Weber, L.; Pepe, F. |
| Abstract: |
The preparation of observations is normally carried out at the telescope
by the visiting observer. In order to help the observer, we propose
several algorithms to automatically optimize the sequence of
targets. The optimization consists of assuring that all the chosen
targets are observable within the given time interval, and to find their
best execution order in terms of the observation quality and the
shortest telescope displacement time. Since an exhaustive search is too
expensive in time, we researched heuristic algorithms, specifically:
Min-Conflict, Non-Sorting Genetic Algorithms and Simulated
Annealing. Multiple metaheuristics are used in parallel to swiftly give
an approximation of the best solution, with all the constraints satis�ed
and the total execution time minimized. The optimization process has a
duration on the order of tens of seconds, allowing for quick
re-adaptation in case of changing atmospheric conditions. The graphical
user interface allows the user to control the parameters of the
optimization process. Therefore, the search can be adjusted in real
time. The module was coded in a way to allow easily the addition of new
constraints, and thus ensure its compatibility with different
instruments. For now, the application runs as a plug-in to the
observation preparation tool called New Short Term Scheduler, which is
used on three spectrographs dedicated to the exoplanets search: HARPS at
the La Silla observatory, HARPS North at the La Palma observatory and
SOPHIE at the Observatoire de Haute-Provence. |
|
|
|
|
|
