Roadmap search based motion planning for MIRADAS probe arms

Sabater, J.; Torres, S.; Garzón, F.; Gómez, J. M.
Referencia bibliográfica

Journal of Astronomical Telescopes, Instruments, and Systems, Volume 4, id. 034001 (2018).

Fecha de publicación:
7
2018
Número de autores
4
Número de autores del IAC
1
Número de citas
1
Número de citas referidas
1
Descripción
MIRADAS is a near-infrared multiobject echelle spectrograph operating at spectral resolution R = 20,000 over the 1 to 2.5 μm bandpass for Gran Telescopio Canarias. It possesses a multiplexing system with 12 cryogenic robotic probe arms, each capable of independently selecting a user-defined target in the instrument field of view. The arms are distributed around a circular bench, becoming a very packed workspace when all of them are in simultaneous operation. Therefore, their motions have to be carefully coordinated. We propose here a motion planning method for the MIRADAS probe arms. Our offline algorithm relies on roadmaps comprising alternative paths, which are discretized in a state-time space. The determination of collision-free trajectories in such space is achieved by means of a graph-search technique. The approach considers the constraints imposed by the particular architecture of the probe arms as well as the limitations of the commercial off-the-shelf motor controllers used in the mechanical design. We test our solution with real science targets and a typical MIRADAS scenario presenting some instances of the two identified collision conflicts that can arise between any pair of probe arms. Experiments show the method is versatile enough to compute trajectories fulfilling the requirements.
Proyectos relacionados
Project Image
Morfología y dinámica de la Vía Láctea

El Proyecto se estructura en dos partes, diferenciadas pero complementarias: morfología y dinámica. El estudio detallado de la morfología de la Vía Láctea pretende proveer una base de datos de distribución estelar en las regiones más alejadas y extintas de nuestra Galaxia, mediante el desarrollo de modelos semiempíricos a partir de la información

Martín
López Corredoira