基于改进概率地图的无人机实时避障研究  

Real time UAV path planning using modified probabilistic roadmap

作　　者：王振华[1] 章卫国[1] 

机构地区：[1]西北工业大学自动化学院,西安710072

出　　处：《计算机工程与应用》2010年第25期220-222,共3页Computer Engineering and Applications

摘　　要：基于空间点采样的概率地图方法能够很好地表示出自由空间的连通性,该方法已在路径规划领域得到了成功的应用。但是,由于在由已得到的采样点基础上构造连通图时,需要检查图的边是否与障碍物发生碰撞,即进行相交检验,限制了概率地图的构造速度,难以满足在实际应用中的实时性要求。针对无人机路径规划问题,以等高线地图作为任务空间,提出了一种新的采样模型,在该模型框架下,依据适当的规则构造临近点集,便可以避免相交检验,提高了路径规划速度。Based on the sampling techniques,the Probabilistic Roadmap Method（PRM） discovers the connectivity of a configoration space very well, and has been successfully implemented in the motion planning field.But, when the sampled points are being connected to each other, collision check is an inevitable step, which is the most time-consuming operation of this method.A new sampling technique is proposed to solve the UAV path planning problem.Using this model, the collision check operation is not necessary.So the path planning operation is speeded up.

关 键 词：概率地图 采样策略 无人机 路径规划 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]