This paper presents a study of fault-detection measurement in redundant systems with an inertial measurement unit on the satellite. In order to deal with system noise, we use a Kalman filter and we employ backpropagation neural networks for fault detection. The simulation results of our measurement compare favourably with those of the SPRT and direct-compare approaches.
Published in:
Intelligent Control and Automation, 2002. Proceedings of the 4th World Congress on
(Volume:2
)
Date of Conference: 2002