A Global Path Planning Algorithm for Fixed-wing UAVs

Abstract

A new approach for solving the global optimal path planning problem to fixed-wing UAVs in multi-threat environments is proposed in this paper, which is mainly based on a natural combination of Genetic Algorithm (GA), Dijkstra searching algorithm, and Artificial Potential Field (APF) approach. First, a Delaunay partition of the flight space is introduced to map the continuous searching space on the Delaunay diagram, and the trajectory encoding methods for GA are designed based on the Delaunay network. Then, a shortest path is established by Dijkstra searching algorithm and the corresponding code is taken as the first population such that a GA could be conducted. Especially by considering flight turning constraints of fixed-wing UAVs, artificial potential field approach is utilized to make the path smooth after each evolution in GA. Finally, a global optimal path is obtained through the suggested algorithm and simulation results validate the effectiveness in both Two-Dimensional (2-D) and Three-Dimensional (3-D) flight environments.