H. Göllinger

Monitoring of sensors aboard an aircraft can be done by comparing the output of several identical instruments (hardware redundancy) or by using models and the inherent redundancy of different sensors (analytical redundancy). Several Kalman filters are used here to detect failures in the measurement of the vertical motion of an aircraft (accelerations, barometric-, GPS-, radioaltitude). Each filter is designed to describe a given sensor failure. The innovations, e.g. the difference between predicted and measured sensor signals are used to find out, which filter fits best to the measured signals. A decision algorithm with variable window size is used to ensure a constant decision quality. The performance of this algorithm is shown using simulated and measured data.

DGLR Symposium, Stuttgart, 1995

Conference Paper

englisch

17,0 x 24 cm, 19 Seiten

DGLR-Bericht, 1995, 1995-03, Aircraft Integrated Monitoring Systems; S.605-623; 1995; Deutsche Gesellschaft für Luft- und Raumfahrt - Lilienthal-Oberth e.V., Bonn

NA

monitoring, kalman filters

Bestellbar

1995