A global path planning method constrained by dynamics, kinematics, and terrain is proposed for Vertical Takeoff and Landing (VTOL) Unmanned Aerial Vehicles (UAVs). Firstly, new constant-altitude waypoints are interpolated above the terrain based on altitude information and terrain resolution. Account for the coverage of airborne sensors over the surface area, a horizontal waypoint interpolation optimization (HOPT) is performed at positions with excessive bending in the horizontal direction. Secondly, for vertical plane optimization (VOPT), i.e., height optimization, a Soft Actor-Critic-based Particle Swarm Optimization (SAC-PSO) is employed to optimize the convergence speed of the method and achieve terrain following (TF). Thirdly, to evaluate the generated paths, a deep residual network (DRSN) is designed to mitigate the impact of optimization failures during the iteration process and improve the stability of the algorithm. Simulation experiments demonstrate the efficiency and path quality of the proposed method, while real-world tasks validate its practicality.
UAV Path Planning With Terrain Constraints for Aerial Scanning
IEEE Transactions on Intelligent Vehicles ; 9 , 1 ; 1189-1203
2024-01-01
4020817 byte
Article (Journal)
Electronic Resource
English
Path planning with variable-fidelity terrain assessment
Tema Archive | 2012
|Fast Path Planning for Fixed-Wing Unmanned Aerial Vehicle with Multiple Constraints
Springer Verlag | 2022
|Optimized Path Planning for Inspection by Unmanned Aerial Vehicles Swarm with Energy Constraints
ArXiv | 2018
|