Unsteady Aerodynamics in Model-Based Wind Estimation from Fixed-Wing Aircraft Motion

This paper examines the value of incorporating unsteady aerodynamics in a model-based wind estimator for a small, fixed-wing aircraft and presents results of an experimental implementation. An extended Kalman filter is designed for two motion models identified from flight data: one that includes unsteady effects and another that does not. An air data unit with wind vanes and a Kiel probe is used to reconstruct the local wind velocity for validation. Analysis of time histories of the wind estimates and the wind estimate error relative to the reconstructed wind velocity shows that the two model-based estimators perform equally well. The power spectral densities of the reconstructed wind and the estimated wind, computed using either estimator, follow Kolmogorov's -5/3 law.