本文为英国莱斯特大学（作者：Rosli bin Omar）的博士论文，共52页。
本文研究基于可视图（VG）方法的无人机（UAV）避障路径规划算法。由于VG使用环境中的所有节点（顶点），因此计算代价很高。与此相反，本文所提出的二维路径规划算法使用所谓的基线（BL）选择相对较少的顶点，因此它们具有较高的计算效率。通过限制BL的长度，进一步提高了算法的计算效率，从而减少了顶点的数目。仿真结果表明，所提出的二维路径规划算法比VG的速度快得多，适合于实时路径规划应用。虽然顶点可以在二维环境中用VG来明确定义，但在三维环境中很难确定它们，因为它们在每个障碍物的边界沿上都是无限的。在本文提出的三维路径规划算法中，顶点是旋转一定角度的平面与障碍物边缘的交点，从而采用所谓的平面旋转方法解决了这一问题。为了保证三维路径规划算法的计算效率，将所提出的二维路径规划算法应用到三维路径规划算法中。此外，还开发了一个基于Matlab的二维和三维路径规划软件包。该软件包简单易用，并具有逐步说明的用户友好性。
This thesis concerns the development ofpath planning algorithms for unmanned aerial vehicles (UAVs) to avoid obstaclesin two- (2D) and three-dimensional (3D) urban environments based on thevisibility graph (VG) method. As VG uses all nodes (vertices) in theenvironments, it is computationally expensive. The proposed 2D path planningalgorithms, on the contrary, select a relatively smaller number of verticesusing the so-called base line (BL), thus they are computationally efficient.The computational efficiency of the proposed algorithms is further improved bylimiting the BL’s length, which results in an even smaller number of vertices.Simulation results have proven that the proposed 2D path planning algorithmsare much faster in comparison with the VG and hence are suitable for real timepath planning applications. While vertices can be explicitly defined in 2Denvironments using VG, it is difficult to determine them in 3D as they areinfinite in number at each obstacle’s border edge. This issue is tackled byusing the so-called plane rotation approach in the proposed 3D path planningalgorithms where the vertices are the intersection points between a planerotated by certain angles and obstacles edges. In order to ensure that the 3Dpath planning algorithms are computationally efficient, the proposed 2D pathplanning algorithms are applied into them. In addition, a software packageusing Matlab for 2D and 3D path planning has also been developed. The packageis designed to be easy to use as well as user-friendly with step-by-stepinstructions.
1 引言
2 路径规划
3 基于可见度线方法的路径规划
4 基于三维可见度线的路径规划
5 用于路径规划的软件工具包
6 结论与未来工作展望
更多精彩文章请关注公众号：
