Simultaneous Search and Tracking of Non-cooperative Mobile Targets Using Multiple UAVs in Uneven Environments

This paper considers the simultaneous search and tracking problem of non-cooperative moving objects in unknown environments for unmanned aerial vehicle (UAV) swarms. The inherent ambiguities in target characteristics increase the complexity of the simultaneous operations, and the necessity for distributed control and adaptability in navigating non-planar terrains also pose significant challenges. This study proposes the target distribution-based cooperative searching and tracking (TDCST) algorithm for multiple UAVs, allowing for executing both search and tracking tasks. The distributed algorithm integrates the target distribution information within the objective function, facilitated by a corresponding information exchange mechanism. Additionally, a zoom camera-based observation (ZCO) model is introduced, considering terrain height variations and utilizing focal length control to mitigate the effects of uneven terrains. This makes the algorithm well suited for searching in uneven terrains. The algorithm ensures a balanced efficiency between searching and tracking tasks, notably enabling the swarm to cover and monitor more targets than the actual number of UAVs, even in scenarios where the exact number of targets remains unknown. Furthermore, the algorithm handles dynamic scenarios where targets appear or disappear unexpectedly. Simulation studies have confirmed the effectiveness and robustness of the proposed algorithm.