一种新的基于切线的路径规划方法  

A New Tangent Based Path Planner

作　　者：吴峰光[1] 奚宏生[1] 

机构地区：[1]中国科学技术大学自动化系,安徽合肥230027

出　　处：《机器人》2004年第3期193-197,共5页Robot

摘　　要：提出了一种崭新的基于切线的路径规划方法 .它在二维离散姿态空间中 ,用障碍物的边界线建立环境模型 ,用凸壳和切线构造局部最短路径 ,复杂地形能被主干线结构有效地分解 .跳跃式扫描技术和按需扩展搜索图的策略使它优于切线图法 .它能非常有效地利用稀疏环境和处理较大的规划空间 ,并且能适应未知和动态的环境 ,这使它成为远距离漫游的理想导航方法 .仿真表明 ,本文算法通常都能得到全局最优路径 ,并且规划速度快、内存需求小 。This paper presents a novel tangent based path planning method. The planner makes use of the obstacle contours to model the environment in a discrete 2D configuration space. Locally shortest paths are constructed by convex hulls and tangents. Complex terrain is effectively decomposed by the Trunk Structure. It makes a major improvement over the tangent graph approach by expanding only necessary parts of the tangent graph and using the Bouncing Scan technique. It is capable of large space, can take full advantage of sparse environments and deal with unknown and dynamic environments, which make it an ideal planner for long range navigation. Simulation results show that it generates global optimal path in most situations, and is very time and space efficient thus suitable for real time applications.

关 键 词：机器人 路径规划 信息规划 环境模型 传感器 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]