Abstract
The estimation from flight data of aerodynamic parameters for vehicle in steady-state conditions, perturbed by an identification manoeuvre, is a well-established technology, whereas system identification from dynamic flight data is a subject of continuous interest. This paper presents a hybrid frequency and time domain technique for identification of vehicle longitudinal aerodynamic model, including the ground effect. Identification is performed in the framework of a multi-step approach, in which, first aerodynamic coefficients are estimated in the frequency domain, using an equation error method; then time domain techniques are applied to identify out of ground effect aerodynamic derivatives and ground effect model parameters. The technique was successfully applied to flight data of an experimental ultra light aircraft. Identification results showed that the proposed method works properly also in the dynamic phases of the flight or when no dedicated identification manoeuvres are executed. Moreover, the identified longitudinal aerodynamic model was used to design the flight control system that successfully performed many autonomous landings.
A Multidimensional Spline-Based Global Nonlinear Aerodynamic Model for the Cessna Citation II
