Research on skeleton reconstruction algorithm in Voronoi graph navigation with improved skeleton extraction
The path planning algorithm based on Voronoi graph will fall into the"trap"of searching the shortest path due to the change of the environment and the invariance of Voronoi graph structure in the course of complex environment navigation,resulting in navigation failure.To solve this problem,a skeleton reconstruction algorithm with improved skeleton extraction in Voronoi graph navigation was proposed.The raster map was preprocessed to generate the global initial skeleton.Combining with the LiDAR observation model to update the cost map,the skeleton reconstruction was assisted by changing the update mode of the cost map to realize the correct navigation of the robot.The experimental results showed that the proposed algorithm could make the skeleton reconstruction"memorized",and improved the navigation robustness of the skeleton algorithm while ensuring real-time navigation.
NETL
NSTL
万方数据
维普
