Flight Test Data for the UAVs

Faculty Contact: Jonathan How

Project funded by the DoD DURIP Program FY2002


The figure shows recent results with two of the UAVs simultaneously flying the same waypoints. By varying the commanded speed we were able to execute a fairly tight formation flying maneuver with the two vehicles.

The fleet of 8 UAVs are flown autonomously using the Cloud Cap commercial autopilot interfaced directly with the planning and task assignment algorithms. Small aircraft (RC-sized trainers) were purposefully chosen to reduce operational complexity while still providing a high degree of flexibility. The UAV testbed has been operated autonomously on numerous occasions - the figure shows the results of one 23 minute autonomous flight. This flight demonstrated our ability to track waypoints and maintain altitude in the presence of moderate wind disturbances. Future flights will include the integration of the receding horizon controller for demonstration of SEAD type missions.

The autopilots have an extensive hardware-in-the-loop simulation capability, which can be used to demonstrate the full suite of coordination and control algorithms prior to flight. The figure shows results from one such demonstration involving 5 UAVs, and a mix of high and low value targets in a complexenvironment. Note that the paths shown contain both the optimal planned trajectories as well as the hardware-in-the-loop autopilot responses (with simulated wind turbulence). For this scenario, high value targets (HVT) require both a strike and subsequent BDA while low value targets (LVT) require strike by a single UAV. The obstacle field divides the region such that the tasks are optimally assigned to two teams of UAVs and one vehicle (UAV 5) is also given the capability to remove the centrally located SAM site. Midway through the mission, UAV one is lost requiring its tasks to be reassigned to other vehicles. UAVs three and four are assigned the remaining tasks and are able to reach their destinations because of the prior SAM site removal. The entire task assignment and trajectory design algorithms are run off-board and are uplinked to the UAV autopilots in real-time exactly as they will be during flight.

Last updated: October 1, 2003