Robust identification of quadrocopter model for control purposes

Abstract

The paper addresses a problem of quadrotor unmanned aerial vehicle (so-called X4-flyer or quadrocopter) utility model identification for control design purposes. To that goal the quadrotor model is assumed to be composed of two abstracted subsystems, namely a rigid body (plant) and four motors equipped with blades (actuators). The model of the former is acquired based on a well-established dynamic equations of motion while the latter is to be identified as a static relationship from laboratory experiments data. Moreover, the actuator model is to account for the on-flight battery power source voltage drop effects. The actuator parameter identification algorithm is kept in a set-membership framework. In addition a mechanism to reduce the conservativeness of the solution is proposed and applied. Numerical illustration of the results is provided.