Engine thrust model determination and analysis using a large operational flight database

Abstract

Different engine thrust models are developed from operational flight data with limited a priori knowledge as part of a novel process for aircraft flight performance model determination. The given big data problem is solved by application of fundamental engineering knowledge and a specific data evaluation strategy. The resulting smart data approach is fundamentally different from existing artificial intelligence methods to solve such big data problems. A linear, a local-linear and a complex nonlinear thrust model are determined on the example of a given large database of operational flights with Airbus A 320neo aircraft. Even with limited information about the actual engine thrust from the available data, the resulting models allow to (well) predict the engine thrust characteristics within the required flight envelope. In addition, a temperature correction is predicted for the thrust model results to further enhance the model's accuracy. Finally, the characteristics of the different thrust model implementations, evaluation results and thrust prediction quality are discussed.