Aiming at the problem that existing UAV indoor exploration path planning methods are too time-consuming,a method for autonomous UAV flight path exploration based on semantic topology and considering the characteristics of the underground environment is proposed.This method is divided into two levels:local path planning and global path planning.In local planning,the environmental point cloud is processed through ground filtering and Delaunay triangulation to obtain high-security local paths and identify intersections.In the global planning,a tree topology is constructed with intersections as nodes and channels as edges,and the re-planning is implemented through a depth-first algorithm to ensure efficient exploration of all channels.Through the simulation verification based on the real environment,the results show that the algorithm can safely complete exploration in the underground environment and significantly shorten the single path planning time.
NETL
NSTL
万方数据
