Estimation of aircraft fuel consumption by modeling flight data from avionics systems

Accurate and economic estimation of aircraft fuel consumption is fundamental for optimizing aviation operations, including emission reduction, flight route planning, and fuel management. Numerous literature presented mathematical models to estimate aircraft fuel consumption but often neglected the challenges of applying those methods in aviation operations. This paper explores a novel strategy to estimate aircraft fuel consumption by modeling flight data from onboard flight data recorder (FDR) and automatic dependent surveillance - broadcast (ADS-B). The Classification and Regression Tree (CART) and Neural Networks (NNs) are adopted for modeling. CART and NN models are developed using FDR data; ADS-B data are used to assess the model performance. The result indicates that the CART model performs better when inputs contain errors and missing values, and the ADS-B data could be used to estimate aircraft fuel consumption as a less-expensive and more convenient strategy compared to the FDR data.

Automated summary

This paper presents a novel strategy to estimate aircraft fuel consumption by modeling flight data from onboard flight data recorder and automatic dependent surveillance-broadcast. The Classification and Regression Tree (CART) model performs better in handling errors and missing values, and the ADS-B data can be used as a less-expensive and more convenient strategy for fuel consumption estimation compared to the FDR data.