Publikationen

DGLR-Publikationsdatenbank - Detailansicht

Titel:

Efficient Sensitivity Calculation for Robust Optimal Control

Autor(en): T. Akman, B. Hosseini, J. Diepolder, B. Grüter, R.J.M. Afonso, M. Gerdts, F. Holzapfel
Zusammenfassung: Due to model and measurement uncertainties, simulation outputs of dynamic systems might not be exact and also subject to uncertainties. Especially in reentry missions unforeseen perturbations and the resulting change of the vehicles behaviour are critical due to high velocities and heat loads. Here, trajectory planning considering uncertainties is of high interest. Therefore, it is important to quantify and reduce the dependency of the system output with respect to uncertainties. In general, reentry trajectories can be calculated by optimal control methods, giving rise to controls such that a certain cost function, e.g. the heat load, is minimized. In this study sensitivities are calculated and minimized in a multi-objective cost function in order to reduce the dependencies of the system regarding uncertain parameters. Two different sensitivity approaches, namely the forward sensitivity and adjoint method are compared. For each of the methods two discretization schemes, first the collocation and second the shooting method, are investigated.
Veranstaltung: Deutscher Luft- und Raumfahrtkongress 2019, Darmstadt
Verlag, Ort: Deutsche Gesellschaft für Luft- und Raumfahrt - Lilienthal-Oberth e.V., Bonn, 2020
Medientyp: Conference Paper
Sprache: englisch
Format: 21,6 x 27,9 cm, 10 Seiten
URN: urn:nbn:de:101:1-2020091411240200307404
DOI:10.25967/490248
Stichworte zum Inhalt: Robust Optimal Control, Forward Sensitivity, Backward Sensitivity, Adjoint, Sensitivity Penalty, Heat Minimal Reentry
Verfügbarkeit: Download - Bitte beachten Sie die Nutzungsbedingungen dieses Dokuments: Copyright protected
Kommentar:
Zitierform:Akman, T.; Hosseini, B.; et al. (2020): Efficient Sensitivity Calculation for Robust Optimal Control. Deutsche Gesellschaft für Luft- und Raumfahrt - Lilienthal-Oberth e.V.. (Text). https://doi.org/10.25967/490248. urn:nbn:de:101:1-2020091411240200307404.
Veröffentlicht am: 14.09.2020