Multi-objective mission planning for UAV swarm search

A phased cooperative mission planning method based on an improved parthenogenetic algorithm and enhanced RRT* algorithm is proposed for UAV swarm search task planning. In the task allocation phase, the multi-target assignment is modeled as a Capacitated Multiple Traveling Salesman Problem. The improved parthenogenetic algorithm incorporates Minkowski distance for path cost estimation and a dynamically adjusted mutation probability strategy to enhance global optimization capability. During trajectory planning, the RRT* algorithm is upgraded through a "semi-circular sampling + artificial potential field" approach, combined with tangent arc transition method to optimize trajectory smoothness and quality. Simulation results demonstrate that compared with benchmark algorithms, the proposed method achieves a 53.2% faster task allocation speed, reduces total trajectory length by 9.2%, and generates trajectories that satisfy UAV kinematic constraints. This provides theoretical support for efficient cooperative mission planning of UAV swarms in complex battlefield scenarios.