TY - GEN
T1 - Data-driven resilience quantification of the US Air transportation network
AU - Chandramouleeswaran, Keshav Ram
AU - Tran, Huy T.
N1 - Publisher Copyright:
© 2018 IEEE.
PY - 2018/5/30
Y1 - 2018/5/30
N2 - This paper presents a data-driven approach for quantifying the resilience of an air transportation network using publicly available data. The methodology relies on a statistical measure, Mahalanobis distance, to detect atypical behavior in the network. This method is applied to parameters of interest from Bureau of Transportation Statistics flight data for the year 2012. The parameters are the total cancellations and average flight delay across all airports. The expected values for these parameters is first established for airports of interest on a given day of the week. The Mahalanobis distance for the network on a given day is then calculated based on these expected values and their variances. Periods with extreme distance values are deemed extreme events. We also separately estimate damages caused by various weather-related events using data from the National Oceanic and Atmospheric Administration to verify results of the distance metric. The distance metric was able to capture impacts of major events like Hurricane Isaac, Hurricane Sandy, and the 2012 North American Blizzard. This method provides a quantitative approach for identifying events that significantly impact the air transportation network. The method can thus support decision-making in designing a more resilient network or air traffic control; for example, by evaluating alternative network topologies or producing training data for disruption impact prediction models.
AB - This paper presents a data-driven approach for quantifying the resilience of an air transportation network using publicly available data. The methodology relies on a statistical measure, Mahalanobis distance, to detect atypical behavior in the network. This method is applied to parameters of interest from Bureau of Transportation Statistics flight data for the year 2012. The parameters are the total cancellations and average flight delay across all airports. The expected values for these parameters is first established for airports of interest on a given day of the week. The Mahalanobis distance for the network on a given day is then calculated based on these expected values and their variances. Periods with extreme distance values are deemed extreme events. We also separately estimate damages caused by various weather-related events using data from the National Oceanic and Atmospheric Administration to verify results of the distance metric. The distance metric was able to capture impacts of major events like Hurricane Isaac, Hurricane Sandy, and the 2012 North American Blizzard. This method provides a quantitative approach for identifying events that significantly impact the air transportation network. The method can thus support decision-making in designing a more resilient network or air traffic control; for example, by evaluating alternative network topologies or producing training data for disruption impact prediction models.
UR - http://www.scopus.com/inward/record.url?scp=85048881028&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85048881028&partnerID=8YFLogxK
U2 - 10.1109/SYSCON.2018.8369602
DO - 10.1109/SYSCON.2018.8369602
M3 - Conference contribution
AN - SCOPUS:85048881028
T3 - 12th Annual IEEE International Systems Conference, SysCon 2018 - Proceedings
SP - 1
EP - 7
BT - 12th Annual IEEE International Systems Conference, SysCon 2018 - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 12th Annual IEEE International Systems Conference, SysCon 2018
Y2 - 24 April 2018 through 26 April 2018
ER -