摘要：AbstractMultirotor unmanned aerial vehicles have become increasingly popular as an automated flying tool in various industrial applications. The requirement of a smooth, accurate and safe flight poses challenges to the design of a suitable controller, especially in typical outdoor environments without precise vehicle state estimations. In this work, a hybrid control approach for such scenarios is presented, which provides exact trajectory tracking performance, good disturbance rejection, and still allows underlying attitude control. A flatness-based state-feedback of the reduced outer dynamics is deduced in order to obtain decoupled linear position and yaw dynamics, whereas the inner attitude dynamics are treated separately and stabilized using conventional linear control as an example. The resulting integrator chains of position and yaw are then controlled using a set of state feedback controllers, and feed-forward action is applied. For analysis, simulations and experiments are carried out in order to compare the proposed hybrid approach to a standard linear control cascade and to a fully linearized flatness-based controller. Results show a similar trajectory tracking performance of the hybrid compared to the full flatness approach, while being significantly more intuitive to implement and robust in typical outdoor environments. The resulting linear input-output dynamics can be extended by any linear control scheme in order to meet the specific requirements of other applications.
