This paper focuses on estimating aircraft mass and thrust setting using a recursive Bayesian method called particle filtering. The method is based on a nonlinear state-space system derived from aircraft point-mass performance models. Using solely ADS-B and Mode-S data, flight states such as position, velocity, and wind speed are collected and used for the estimation. An important aspect of particle filtering is noise modeling. Four noise models are proposed in this paper based on the native ADS-B Navigation Accuracy Category (NAC) parameters. Simulations, experiments, and validation, based on a number of flights are carried out to test the theory. As a result, convergence of the estimation can usually be obtained within 30 seconds for any climbing flight. The method proposed in this paper not only provides final estimates, but also defines the limits of noise above which estimation of mass and thrust becomes impossible. When validated with a dataset consisting of the measured true mass and thrust of 50 Cessna Citation II flights, the stochastic recursive Bayesian approach proposed in this paper yields a mean absolute error of 4.6%.
Original languageEnglish
Title of host publication2018 International Conference on Research in Air Transportation
Subtitle of host publicationBarcelona, Spain, 2018
Number of pages8
Publication statusPublished - 2018
EventICRAT 2018: 2018 International Conference on Research in Air Transportation - Castelldefels, Barcelona, Spain
Duration: 26 Jun 201829 Jun 2018
Conference number: 8


ConferenceICRAT 2018: 2018 International Conference on Research in Air Transportation
Abbreviated titleICRAT 2018
Internet address

    Research areas

  • aircraft, state estimation, point-mass model, measurement noise, particle filter, Bayesian estimation

ID: 47548834