Aerial traffic planning for autonomous multi-rotor in a dynamic known environment based on a Decomposition-Coordination method

This paper deals with the problem of aerial traffic planning for autonomous Unmanned Aerial Vehicle (UAV) and deploys a successful Decomposition-Coordination Method (DCM) to provide the optimal solution. Our purpose is to make the UAV's flight smoother and avoid the dynamic obstacles. We work under the assumption that the environment is known and the obstacles positions can be identified in real-time by a deciding entity. We choose an initial position, for which we plan a first safe and optimal path which remains valid until a moving obstacle crosses it. To avoid the collisions, the deciding entity provides a new safe path, allowing the UAV enough freedom to reach the desired position safely. For this matter, we choose a nonlinear model that describes veraciously the rotational and translational dynamics of the UAV, we then use the DCM to compute the optimal control, thus enabling the UAV to navigate toward the safe path. This approach consists of breaking down the nonlinear system into several interconnected subsystems which allows the non-linearity to be treated at a local level, we then use the Lagrange multipliers to achieve coordination. Consequently, parallel processing minimizes remarkably the computation time which offers a fast solution to the traffic problem. to demonstrate the convergence and the stability of the method we make use of two theorems. The numerical application and the simulation presented confirms the efficiency of this technique.