Adaptive strategies for collaborative work with scale quad-rotors

The purpose of this paper is to present strategies for the control of movement of rigid bodies via force actuators, possibly redundant. After a nonlinear feedback linealization of the considered dynamic models and the application of a suitable controller, an adaptive neural network based control component is incorporated in order to cope with modeling errors and disturbance rejection. An online sequential quadratic programing optimization with equality and inequality constraints assures an adequate configuration of actuator forces. Application to collaborative work in the transportation of a rigid body using a squadron of scale quad-rotors is studied.