A Non Linear Model Predictive Controller (NMPC) for the realization of an autonomous take-off procedure of a rigid wing Airborne Wind Energy (AWE) system is proposed. The addressed take-off approach employs a linear motion system to accelerate the aircraft to take-off speed, a winch which is connected to the aircraft with a tether, and finally on-board propellers to drive the aircraft to operational altitude. A comparison between the performance of the NMPC with the Linear Quadratic Regulator (LQR) is made to determine whether it is necessary to use advanced non-linear controllers in the considered takeoff procedure. Simulations results demonstrate that the NMPC outperforms the LQR in terms of pitch angle and velocity tracking as well as time to reach a desired minimum altitude. Additionally, a comparative power analysis was conducted and results indicate that the NMPC controller is more energy efficient since it is able to drive the aircraft to a larger altitude with smaller power losses.

This paper addresses the challenges associated with formation control of quadrotors. The collaboration between quadrotors necessitates communication, introducing additional complexities related to the network, specifically time delays and network switching. Although several algorithms have been proposed in prior research to handle these challenges separately, none have tackled both time delays and network switching concurrently. Our objective is to design a control law that handles both time delays and network switching concurrently. Initially, we use a control strategy consisting of three parts: 1) formation controller, 2) reference-angle generator and 3) attitude controller, for the simple case with no network imperfections. Subsequently, we adapt this strategy to accommodate both time delays and network switching, thereby providing a solution to the formation control problem for quadrotors under network imperfections.

This paper studies formation control problem for linear heterogenous multi-agent systems operating in the presence of network imperfections, specifically arbitrary switching and arbitrary time delay. Currently, there exists various control structures for handling this problem. One class of control structure that stands out, is a control law that uses only neighbours’ output information instead of the full state to reduce communication burden. However, existing methods, such as those employing Linear Matrix Inequalities (LMIs), overlook network switching. Another approach involves pole placement for individual agents. However, the proof of stability neglects arbitrary time delays, and the placement of poles is often constrained. The main contribution of this work is to present a novel proof for the stability of heterogeneous systems of linear agents, even in the presence of arbitrary switching and arbitrary time delays, using the concept of positive impulse systems.