Traffic optimization for UAV using a method of decomposition-Coordination

This paper presents an innovative solution for the traffic optimization problem, for an autonomous quadcopter Unmanned Aerial Vehicle (UAV). The overall purpose of this work is to allow the flying robot to navigate safely to the desired position, in a dynamic environment. We consider a supervisory unit receiving all the necessary information on the dynamic obstacles, and the moment they may appear on the road so that it set a first optimal /collision-free trajectory. The UAV navigate accordingly until a dynamic obstacle comes across, the supervisory unit provides then a safe trajectory that the quadcopter should reach to avoid the collision. To solve this nonlinear problem, we propose a Decomposition-Coordination method (DC) that enables the autonomous UAV to compute an optimal control to reach the next planned path in real-time. The principal of this method is based on a decomposition of the nonlinear system into a set of interconnected subsystems, to treat the non-linearity locally, we make use then of the Lagrange multipliers to achieve the coordination. This parallel processing minimizes enormously the computation time and offers a faster re-activity to the UAV. The convergence and stability of the method are dully proven, and the numerical application along with the simulation results are presented to affirm the potential of our approach.