TY - GEN
T1 - Efficient Estimation of Probability of Exceeding Performance Limits for Automatic Landing Systems using Subset Simulation
AU - Mishra, Chinmaya
AU - Schwaiger, Florian
AU - Blum, Christopher
AU - Kleser, Marc Andreas
AU - Krammer, Christoph
AU - Holzapfel, Florian
AU - States, True
N1 - Publisher Copyright:
© 2022, American Institute of Aeronautics and Astronautics Inc. All rights reserved.
PY - 2022
Y1 - 2022
N2 - The certification of automatic landing systems requires demonstration of compliance with the requirements outlined by aviation authorities. In recent years, the certification authorities have shifted from prescriptive requirements to performance-based requirements. Compliance may be shown with a combination of flight tests and simulations that evaluate the performance of the landing. The requirements consist of performance criteria and limits with probability thresholds respectively. The probability of exceedance of the limits must be below the threshold. Such probability thresholds are very low (of the order 10-5 to 10-8 ) and require a significantly large number of samples to be realized by naive Direct Monte Carlo simulations. The increased number of samples leads to an increase in computational load which results in higher development costs. For cases which require estimation of low probabilities, a method called subset simulation has been demonstrated to be efficient in estimating low probabilities using significantly fewer samples than Direct Monte Carlo. Subset simulation uses a combination of Direct Monte Carlo and Markov chain Monte Carlo to estimate low probabilities as a product of larger probabilities. This paper presents the application of the Markov chain Monte Carlo toolchain based on subset simulation developed within the Institute of Flight Systems Dynamics at the Technical University of Munich to estimate the low probability of exceeding performance limits for an automatic landing system developed for the institute’s DA42 research aircraft. An example automatic landing scenario consisting of a six-degree-of-freedom flight dynamic model in closed-loop control with automatic flight guidance and control algorithms, an integrated navigation system and sensor models is used to estimate the probability of exceeding performance-based requirements outlined in CS-AWO.
AB - The certification of automatic landing systems requires demonstration of compliance with the requirements outlined by aviation authorities. In recent years, the certification authorities have shifted from prescriptive requirements to performance-based requirements. Compliance may be shown with a combination of flight tests and simulations that evaluate the performance of the landing. The requirements consist of performance criteria and limits with probability thresholds respectively. The probability of exceedance of the limits must be below the threshold. Such probability thresholds are very low (of the order 10-5 to 10-8 ) and require a significantly large number of samples to be realized by naive Direct Monte Carlo simulations. The increased number of samples leads to an increase in computational load which results in higher development costs. For cases which require estimation of low probabilities, a method called subset simulation has been demonstrated to be efficient in estimating low probabilities using significantly fewer samples than Direct Monte Carlo. Subset simulation uses a combination of Direct Monte Carlo and Markov chain Monte Carlo to estimate low probabilities as a product of larger probabilities. This paper presents the application of the Markov chain Monte Carlo toolchain based on subset simulation developed within the Institute of Flight Systems Dynamics at the Technical University of Munich to estimate the low probability of exceeding performance limits for an automatic landing system developed for the institute’s DA42 research aircraft. An example automatic landing scenario consisting of a six-degree-of-freedom flight dynamic model in closed-loop control with automatic flight guidance and control algorithms, an integrated navigation system and sensor models is used to estimate the probability of exceeding performance-based requirements outlined in CS-AWO.
UR - http://www.scopus.com/inward/record.url?scp=85123891981&partnerID=8YFLogxK
U2 - 10.2514/6.2022-1894
DO - 10.2514/6.2022-1894
M3 - Conference contribution
AN - SCOPUS:85123891981
SN - 9781624106316
T3 - AIAA Science and Technology Forum and Exposition, AIAA SciTech Forum 2022
BT - AIAA SciTech Forum 2022
PB - American Institute of Aeronautics and Astronautics Inc, AIAA
T2 - AIAA Science and Technology Forum and Exposition, AIAA SciTech Forum 2022
Y2 - 3 January 2022 through 7 January 2022
ER -