Flight performance analysis with data-driven mission parameterization: mapping flight operational data to aircraft performance analysis

Abstract

This paper develops a hybrid data-driven physics-based model, which maps flight operational data to aircraft performance analysis, in order to estimate aircraft fuel consumption and flight condition information. The hybrid model is an adaptation of the enhanced high-fidelity flight performance analysis, with adjustments made for mission parameters. Through the proposed mission parameterization process, nominal parameters are substituted by the extracted mission and speed profiles. In other words, real flight operations are integrated into the physics-based analysis. The versatility of the hybrid model is demonstrated in three cases (short-, medium-, and long-haul flight sectors), which have very different characteristics. It is also validated against the corresponding fuel consumption data from an airline (with root mean square errors less than 3%) and shows great improvements relative to the pure physics-based model.