Load transportation by individual and cooperative quadrotors using adaptive and sliding mode control

Abstract

This thesis presents a practical study on the topic of load transportation using quadrotors. In specific, along the present works we treat the transported payload as an unmodeled disturbance, and different approaches are tested to deal with this kind of system. The difficulties involving individual and cooperative load transportation are analyzed, as well as payload orientation, in the case of cooperative transportation. For a broad understanding of the subject, a comprehensive state-of-the-art review is firstly presented. In the following, we proposed different frameworks to deal with the model uncertainties and disturbances caused by the payload, or by the dynamic effects caused by one vehicle on the other, in the case of cooperative transportation. We proposed these frameworks considering two major branches in engineering to improve robustness: model reference adaptive control and sliding mode control. To validate the proposals, we tested our algorithms in adverse conditions, under wind-like disturbances and outdoors, for instance, and in transportation missions far from quasi-static motion. As a result, the performance during the experiments demonstrated the effectiveness in disturbance rejection for the here proposed robust adaptive controller and adaptive sliding mode controllers, allowing safe and agile transportation when considering a single quadrotor or a team of two quadrotors for the task. For the sake of completeness, a qualitative analysis of the practical results obtained is also provided, emphasizing the relevance of this work in face of the state-of-the-art.