An intelligent task assignment algorithm for UAVs cluster for fast-moving targets

With its high flexibility, wide adaptability and controllable economy, UAVs cluster has more and more extensive application potential, and has been highly concerned at home and abroad. Task assignment is the top-level planning of UAVs cluster application, which is a comprehensive scheduling according to the requirements of mission environment situation, mission requirements, and own characteristics, so as to establish a reasonable mapping relationship between UAV and task, and maintain a reasonable cooperative relationship between aircraft. In this paper, dynamic model-based modeling and Kalman filter trajectory prediction for fast-moving targets are adopted to improve the accuracy and prediction ability of UAVs cluster on target trajectory. Then, an improved particle swarm optimization algorithm is proposed, which increases the global iterative optimization ability of the initial particle swarm by using linear decreasing parameter Settings and large inertia weights and learning factors as the beginning. And it ends with a small inertia weight and learning factor, which strengthens the final particle swarm to jump out of the local optimal solution. Intelligent cooperation and task allocation among UAVs cluster are realized, thus improving the efficiency and flexibility of the whole system. Finally, through task assignment, the UAVs cluster can track each fast-moving target autonomously.