A methodology for the identification of inertial properties of small size UAVs

This paper illustrates a procedure for estimating the inertial properties of small size aerial vehicles. An identification algorithm has been developed that, starting from experimental data, estimates the parameters of a physical model describing the pendular motion of a generic rigid body. The attitude time histories of a structure (\cage"), carrying the object whose inertial properties have to be evaluated, are the experimental data obtained through a measurement unit attached to the cage itself. The cage, designed in order to facilitate the assembly issues, is put in pendular motion thanks to a needle shaped pivot, placed to the cage top and leaning against a beam. Before proceeding to the identification of the aerial vehicle inertial properties, several tests have been performed to evaluate the performance of the algorithm. A preliminary effectiveness of the algorithm has been assessed via simulation environment, by artificially creating \virtual" time histories. Afterwards, the algorithm has been validated experimentally by loading the cage with a proof mass of known inertial characteristics. During these experimental tests, specific attention has been focused on the effect of the cage initial attitude on the inertial properties estimate. After this algorithm test phase, the developed methodology has been applied to a small rotary-wing UAV in order to evaluate its inertial properties.