D. Martens

The need of system monitoring has dramatically increased along with the use of more and more complex technical aircraft equipment. Monitoring algorithms often rely on statistical analysis thus being time and memory consuming and difficult to certify. At the Institute of Flight Guidance and Control of the Technical University of Braunschweig some feasibility studies as well as first simulations in the field of sensor monitoring with artificial neural networks have been performed. Due to the flexibility and robustness of neural networks as well as the possibility of their hardware implementation they could be an alternative to algorithmic systems of future aircraft. First promising results are presented in this paper. They include monitoring of inertial sensors as well as satellite based navigation systems.

DGLR Symposium, Stuttgart, 1995

Conference Paper

englisch

17,0 x 24 cm, 18 Seiten

DGLR-Bericht, 1995, 1995-03, Aircraft Integrated Monitoring Systems; S.675-692; 1995; Deutsche Gesellschaft für Luft- und Raumfahrt - Lilienthal-Oberth e.V., Bonn

NA

monitoring, neural nets

Bestellbar

1995