Piecewise Polynomial Model Identification using Constrained Least Squares for UAS Stall

Vincent Guibert (1, 2) , Jean-Philippe Condomines (1) , Mathieu Brunot (2) , Murat Bronz (1)

1 ENAC - Ecole Nationale de l'Aviation Civile
2 DTIS, ONERA, Université de Toulouse [Toulouse]

In this paper, we propose a new piecewise polynomial model (PwPM) for the modelling of an Unmanned Aircraft System (UAS) aerodynamic coefficients over a wide flight envelope, where they are characterized by nonlinear and hysteresis phenomena. An associated identification method using Constrained Least Squares (CLS) is then presented and successfully applied to experimental data obtained from wind tunnel measurements. The resulting fitting is finally compared with the similar piecewise models PWPFIT and AERODAS, showing a greater precision in its modelling, while maintaining their relative simplicity and computation performance.

Mots clés: Flight dynamics identification; Piecewise polynomial; Constrained least squares; Unmanned aircraft system; Stall modelling

Domaines: Automatique / Robotique

Vincent Guibert, Jean-Philippe Condomines, Mathieu Brunot, Murat Bronz. Piecewise Polynomial Model Identification using Constrained Least Squares for UAS Stall. 19th IFAC Symposium on System Identification, Jul 2021, Padova (virtual), Italy. ⟨10.1016/j.ifacol.2021.08.408⟩. ⟨hal-03286059⟩