Design and simulation of a star tracker for the Aramis nanosatellite

Abstract

Star Trackers aredevicesthat provide higheraccuracy than other attitude sensors with the added benefits of 3-axis attitude determination. Nevertheless, Star Trackers are frequently heavy, complex and costly systems that can not be adopted by small satellites such asthe ARAMIS from Politecnico di Torino, which needs high-accuracy attitude determination to cover te requirementes of certian types of payload.

A preliminary desing of a low-mass, low-cost, low-power and coarse accuracy Star Tracker is proposed to satisfy the requirements of the ARAMIS spacecraft. Diferent available algorithms for identifying the presence of single stas on the imager plane are analyzed, as well as those for pattern recognition necessary to ultimately measura the spacecraft attitude. One set of such image processing and pattern recognition algorithms are chosen for use on board Aramis. Subsequently, they are tested with the experimental use of the 3D open source planetarium Celestia, while a parallel test of the image processing algorithms is performed on real star field imagery to confirm their capabilities with real-word data.

A scheme is proposed to reduce the amount of false results thanks to the use of attitude approximations coming from other sensor, through the homogeneous segmentation of the celestial sphere.