TY - EJOU
AU - DeSouza, L.C.G.
TI - Application of Kalman Filter and H^{∞} Methodologies to Estimate Attitude of a Satellite Control System Simulator
T2 - The International Conference on Computational \& Experimental Engineering and Sciences
PY - 2008
VL - 5
IS - 3
SN - 1933-2815
AB - Satellite Attitude Control System usually does not have all the states available for feedback; therefore, full state estimation by any sort of filtering methodology becomes essential. A good estimation algorithm must filter out the undesirable noise from the measurement signal. Kalman Filter (KF) technique is a traditional procedure to estimate the states of a linear system that are not always measured directly by the sensors, minimizing the variance of the estimated error. However, the KF is not fully robustness proven in face of unpredictable noise sources and it is not always able to minimize the error under severe conditions. In that case, the H^{∞} filter method is an alternative when robustness is at stake. The H^{∞} filter is less known and less commonly applied than the Kalman filter for state estimation, but it presents advantages that make it more effective in some situations. This paper presents the application and comparison between the conventional Kalman filter and the H^{∞} technique for estimating attitude of a Satellite Attitude Control System Simulator, which has a reaction wheel as actuator and a gyroscope as sensor. Both filters performance are investigated considering noise variation due to uncertainties in the plant and sensors.
KW - Estimation
KW - control system
KW - satellite simulator
DO - 10.3970/icces.2008.005.163