Cooperative UAV Tracking Under Urban Occlusions and Airspace Limitations

A centralized approach for multi UAV cooperative motion planning, for tracking a predictable ground moving target in urban environments with airspace limitations, using gimballed or body flxed sensors, is presented. Automating this task is motivated by the expected reduction in operators’ workload and performance improvement. To ensure ∞yable trajectories, adequate performance, and safety, the UAVs’ dynamics, occlusions and airspace constrains must all be incorporated into the problem’s formulation. The solution strategy involves determining visibility, sensor coverage, and restricted regions in the calculated horizon using either a priori or operator provided information on the urban terrain and target trajectory. The tracking task is then casted as a centralized optimization motion planning problem, in which the cost function is associated with the UAVs’ positions relative to the visibility and restricted regions, and the target’s position relative to the sensor coverage region. A computationally parsimonious stochastic search method (genetic algorithm) is proposed for solving the resulting optimization problem. The algorithm was implemented in a high fldelity simulation test-bed using a visual database of an actual city. The viability of using the algorithm is shown using a Monte Carlo study.