Unmanned Aerial Vehicles – ICSL

Unmanned Aerial Vehicle Group

The unmanned aerial vehicle (UAV) group of the ICSL develops algorithms that add to the autonomy of both single and multiple UAV systems. The group is currently studying the following topics:

Differential Games — A multivehicle controller is currently being developed to guide a swarm of unmanned aerial vehicles (UAVs) in adversarial situations. The strategies used to guide the UAVs are derived online by solving multiple simple two-player differential games, and then combine the results to generate a complete strategy for the entire swarm. Games Intro.JPG
Visual Tracking — The visual tracking problem is a state estimation problem. However, due to inherent clutter in the image, conventional estimation techniques, such as the Kalman filter are not acceptable. Therefore, an adaptive particle filter is used. The adaptation allows the filter to find the appropriate model and filter parameters to not only estimate the position of the target, but also operate in real-time. PfSample.jpg
Coverage Planner — The multiple vehicle coverage planner (MVCP) generates efficient routes for multiple UAVs performing a coverage task. Spanning Trees are used as the basis for the route planning algorithm. The resulting system is computationally efficient and guarantees full coverage of the region of interest. Coverage2.gif