Robust formation reconfiguration control methodology for unmanned aerial vehicles

Abstract

Formation flying of Unmanned Aerial Vehicles (UAVs) has gain a lot of interest due to its many potential advantages. Flying in formation allows wider sensing coverage area and in effect, this leads to improved surveillance capability and enhanced situational awareness. Also flying in formation eases coordination and data fusion. This work addresses the issue of reconfiguration during a formation flight. It is an important issue, because at some points during its flight, the formation may need to be changed due to mission demands. Previously, the approach of such problem is by considering the reconfiguration as a part of mission planning. However, in order to provide a collision-free trajectories, the algorithm involved is usually time consuming and often computationally expensive.

This work focuses on the reconfiguration control for non-hover-capable UAVs. The current premise is for the UAVs to assume their final target states within a specified time interval while avoiding collisions with one another or with an obstacle during the process. To achieve the mission, a reference trajectory is generated based on the two-point boundary values (TPBV) optimization of a double integrator. Sliding controller is then used to track the generated reference trajectory for its robustness. To avoid collisions, the area within the UAV's sensor detection range is modeled as a potential-like function that feeds directive control signals which varies with the distance between the UAV and the obstacles (could be stationary or moving) nearby. Such controller is shown to be stable in the sense of Lyapunov. Some simulations are carried out to assess the performance of the reconfiguration control scheme. It is found that provided sufficient safety margin exists between the moment of obstacle detection to the minimum safety radius of the UAV, the proposed control scheme gives a satisfactory performance for a collision-free reconfiguration.