Data-Driven Optimal Control for Image-Based Cooperative Target Tracking of Multiple UAVs

This paper implements an optimal controller for multi-UAV cooperative target tracking based on image data. To enable UAVs to perform more complex functions, a multi-UAV cooperative control approach is adopted. The UAV equipped with a camera acts as the leader of the formation control, performing target tracking using image-based visual servoing (IBVS), while the other UAVs serve as followers in the formation. Due to communication constraints among UAVs, the leader cannot directly communicate with all followers. Thus, a distributed estimator is developed to determine the target positions of individual UAVs. Moreover, acquiring precise dynamic model for each UAV in real-world applications is often difficult. To address this, a data-driven strategy iteration method is utilized to derive the optimal solution when dynamic models are unavailable. A series of simulations were performed to validate the reliability of this approach.