DocumentCode
2568975
Title
Decomposition methods for optimized collision avoidance with multiple threats
Author
Chryssanthacopoulos, James ; Kochenderfer, Mykel
Author_Institution
MIT Lincoln Laboratory
fYear
2011
fDate
16-20 Oct. 2011
Firstpage
1
Lastpage
16
Abstract
• Decomposition methods for solving the multiple-threat Markov decision process leverage the solution to the single-threat problem • Decomposition methods outperformed the current version of the Traffic Alert and Collision Avoidance System in realistic three-aircraft simulations • Utility fusion methods, particularly max-sum, performed better than command arbitration methods by fusing utilities associated with different intruders • Utility fusion methods provide safety equivalent to an example global method with fewer strengthenings
fLanguage
English
Publisher
ieee
Conference_Titel
Digital Avionics Systems Conference (DASC), 2011 IEEE/AIAA 30th
Conference_Location
Seattle, WA, USA
ISSN
2155-7195
Print_ISBN
978-1-61284-797-9
Type
conf
DOI
10.1109/DASC.2011.6096170
Filename
6096170
Link To Document