3D PRM based real-time path planning for UAV in complex environment
Fei Yan, Yan Zhuang, Jizhong Xiao
Abstract
Fei Yan, Yan Zhuang, Jizhong Xiao
Abstract
This paper presents a real-time path planning algorithm for unmanned aerial vehicle (UAV) in complex 3D environments. In this study, the work space is represented by free voxels, which have enough space tolerance to satisfy the safety needs of the UVA. In order to obtain the connectivity of free voxels effectively, the environment is divided into bounding box array and the connectivity evaluation is carried out in each bounding box. A modified Probabilistic Roadmap Method (PRM) is introduced by random sampling in bounding box array to ensure a more evenly distribution of roadmap nodes in 3D space. Based on the voxel connectivity, the selected nodes compose a roadmap, which is applied for path searching by A* algorithm for feasible path. Experimental results show that the proposed path planning approach can significantly reduce the computation time than traditional PRM method.
OpenAlex reports 24 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This paper presents a real-time path planning algorithm for unmanned aerial vehicle (UAV) in complex 3D environments. In this study, the work space is represented by free voxels, which have enough space tolerance to satisfy the safety needs of the UVA. In order to obtain the connectivity of free voxels effectively, the environment is divided into bounding box array and the connectivity evaluation is carried out in each bounding box. A modified Probabilistic Roadmap Method (PRM) is introduced by random sampling in bounding box array to ensure a more evenly distribution of roadmap nodes in 3D space. Based on the voxel connectivity, the selected nodes compose a roadmap, which is applied for path searching by A* algorithm for feasible path. Experimental results show that the proposed path planning approach can significantly reduce the computation time than traditional PRM method.
Key concepts: Minimum bounding box, Motion planning, Voxel, Probabilistic roadmap, Computer science, Computation, Bounding overwatch, Path (computing)