Attitude determination for the three-axis spacecraft simulator (TASS) by application of particle filtering techniques

Abstract

The accurate determination of spacecraft attitude has always been a critical issue in many applications. The presence of imperfect sensors introduces errors in the system and affects the outcome of the mission. One of the most significant sensors is the rate gyroscope. Particularly, the rate gyros are known to degrade with time, introducing random noise and bias. This calls for estimation algorithms which process the measured data in order to reduce the effects of the disturbances to a minimum. This research presents an approach which takes full advantage on the nonlinear dynamics and possibly non-Gaussian disturbances. It is based on recent work involving particle filters, where the probability density functions are approximated by a relatively large number of parameters. It is shown that accurate attitude estimation can be obtained with a manageable number of particles.