REAL-TIME STEALTH CORRIDOR PATH PLANNING FOR FLEETS OF UNMANNED AERIAL VEHICLES IN LOW-ALTITUDE PENETRATION

An improved niched genetic algorithm (NGA) is presented for stealth corridor path real-time planning for heterogeneous unmanned aerial vehicles (UAVs) in three-dimensional (3D) dynamic environments. Firstly, the path planning problem is formulated. 3D corridor paths are introduced to meet the kinematic constraints of a UAVs fleet. Secondly, NGA is improved by merging the neighbourhood field operator and the rollback operator. In addition, the crowding strategy is employed to generate K coverage paths in the searching space. To balance converging rate and optimality, K near-optimal paths are used as the final solutions. Simulation results demonstrate that the proposed algorithm converges more rapidly in terms of time and high-quality flight corridors for UAVs fleet can be obtained in real time.