Probabilistic Characterization of Operational Uncertainties in Transport Aircraft using OpenSky

Abstract

The aerodynamic design of transonic wings is already a mature field, and the use of aerodynamic shape optimization is a well-established discipline in industrial setting. Aircraft manufacturers design configurations by considering a representative but limited set of flight conditions. In practice, airlines do not always fly at the conditions they were designed to operate. Flight altitude, airspeed and aircraft weight are affected by operational requirements and environmental uncertainties. As a result, aircraft altitude, Mach number and lift coefficient, three of the most important parameters when performing aerodynamic design, can not be treated as single deterministic values in the design process. A full probabilistic approach is required to better characterize the real performance of the aircraft. However, there is a lack of aircraft operational data necessary to characterize uncertainty sources in flight. The objective of this paper is the characterization and quantification of operational uncertainty sources based on aircraft surveillance data. The definition of these uncertainties will be essential for the robust design of the next generation of commercial aircraft. To understand the variability in operating conditions of a representative aircraft fleet, surveillance data from the OpenSky network is gathered. The Mach number is directly obtained from the BDS-60 codes, while the altitude is provided by the ADS-B. The lift coefficient of the aircraft at each instant is roughly estimated according to the Breguet equation and the initial and final fuel weights. These are determined by the distance between departure and arriving airports. After the Mach, lift coefficient and altitude are obtained for each individual flight, they are filtered for cruise conditions (level flight). A Kernel Density Estimation is used to obtain the full probability distribution function. This methodology enables the accurate characterization of operational uncertainties that will be required for the aerodynamic robust design of the next generation of aircraft. The design will be tailored to the airliners operations. This framework can also be used by designers and operators to understand how aircraft are operated in reality.