Nonlinear Model Identification of a Small-Scale Unmanned Rotorcraft

This paper illustrates a non-linear model for the dynamic simulation of small-scale helicopter and the activities carried out for the identification of the unknown parameters. Recently, the University of Pisa has undertaken activities aiming to the development of small-scale unmanned systems starting from small commercial RC model helicopters. In this context, the simulation plays a crucial role, for both performance characterisation and flight control laws design, and accurate mathematical models of such vehicles are required. The simulation model of the helicopter dynamics has been developed in Matlab/Simulink environment by minimising the number of parameters to be identified. To this end, a detailed mathematical model for the actuation system of the blades collective and cyclic pitch has been developed, which takes into account the Bell-Hiller mixer effects. The model relating the three servo-actuators rotations to the blades pitch is based on complex nonlinear equations, while the rotor aerodynamics has been modelled with the blade momentum theory, coupled to the inflow Glauert predictions. Concerning the identification process, a database has been collected by carrying out specific flight tests with the helicopter equipped with a GPS, inertial sensors and a data-acquisition system. Contrary to the literature examples, the proposed identification method operates on a nonlinear model in the time domain, rather than on linear models and (often) in the frequency domain. The proposed approach leads to an identification based on a minimum set of unknown parameter and assures a satisfactory matching between simulation predictions and experimental data on the whole flight envelope.