Distributed cooperative search method for multi-UAV with unstable communications

In search-attack and search-rescue tasks performed by multiple unmanned aerial vehicles (multi-UAV), cooperative search plays an important part. The majority of approaches in use today assume that the UAV swarm's communication network is complete. However, these links are susceptible to environmental changes or adversary interference. Aiming at cooperative search in search-attack and search-rescue tasks, we propose a distributed cooperative search method for multi-UAV with unstable communications (DCS-UC), which is developed based on ant colony optimization (ACO). The proposed method presents three algorithms to enable the UAV swarm to conduct online cooperative search efficiently and safely. First, a pheromone matrix consensus update technique for usage in ACO is designed under switching and connected topology graphs. This approach can help each UAV's pheromone matrix achieve consensus, improving the UAV swarm's cooperative efficiency. Then, a position consensus update algorithm is presented where the position vector, including the positions of all the UAVs in each UAV, can achieve consensus under switching and connected topology graphs. Finally, a collision avoidance algorithm is developed using the determined consistent positions of all the UAVs to address the issue of collision avoidance for all the UAVs. This approach enables the UAV swarm to conduct cooperative search safely. Results from extensive physical simulations performed in Gazebo confirm the benefits of the proposed multi-UAV cooperative search method.