AAS 96-173

EULER-q ALGORITHM FOR ATTITUDE DETERMINATION FROM VECTOR OBSERVATIONS

D. Mortari, Universita degli Studi di Roma "La Sapienza"

Abstract

In this paper a new optimal attitude criterion is proposed and an estimation algorithm, EULER-q, is derived from it. The solution is provided in a closed-form and the singular case is analyzed. Then an improvement (QUEST-2) of the QUEST algorithm, is presented. It provides the eigenvalues in a closed-form and computes the quaterion by a generalized cross product in a 4D-space. The parameter describing the rotational error is then analyzed and found. Accuracy tests show the new optimal criterion having an average very small precision loss, which decreases as the number of observed directions increases, and speed tests show EULER-q and QUEST-2 as the presently fastest optimal estimation algorithms.