Robust, Terrain-Aided Landing Navigation through Decentralized Fusion and Random Finite Sets
This paper explores the use of simultaneous localization and mapping with random finites sets to process terrain camera data and augment traditional Kalman filter-based navigation tools. Enabled by new theoretical developments, the presented methods employ both consider parameters and square-root factors of covariance in the interest of producing a practical, numerically stable navigation algorithm. Additionally, a model is presented for a data-driven learning procedure that permits observations of previously unexplored terrain to be instantiated into the filter and improve subsequent navigation estimates. The proposed techniques are evaluated in a simulated lunar descent scenario, and a Monte Carlo analysis confirms the statistical consistency of the described methods.
J. S. McCabe and K. J. DeMars, "Robust, Terrain-Aided Landing Navigation through Decentralized Fusion and Random Finite Sets," Proceedings of the AIAA Guidance, Navigation, and Control Conference (2018, Kissimmee, FL), no. 210039, American Institute of Aeronautics and Astronautics (AIAA), Jan 2018.
The definitive version is available at https://doi.org/10.2514/6.2018-1332
AIAA Guidance, Navigation, and Control Conference, 2018 (2018: Jan. 8-12, Kissimmee, FL)
Mechanical and Aerospace Engineering
Keywords and Phrases
Aviation; Kalman filters; Landforms; Monte Carlo methods; Program processors, Decentralized fusions; Learning procedures; Monte carlo analysis; Navigation algorithms; Simultaneous localization and mapping; Statistical consistencies; Theoretical development; Traditional Kalman filters, Navigation
International Standard Book Number (ISBN)
Article - Conference proceedings
© 2018 American Institute of Aeronautics and Astronautics (AIAA), All rights reserved.
01 Jan 2018