面向对象的自主车越野路径规划的设计和实现  

DESIGN AND IMPLEMENTATION OF AN OBJECT-ORIENTED CROSS-COUNTRY PATH PLANNING FOR AN AUTONOMOUS LAND VEHICLE

作　　者：翁妙凤[1] 潘峻[1] 

机构地区：[1]华东船舶工业学院

出　　处：《计算机研究与发展》1996年第7期533-540,共8页Journal of Computer Research and Development

摘　　要：本文介绍了一种使用启发式估值函数进行二次搜索的路径规划方法，设计和实现了自主车的越野路径规划。首先是在领域式规划空间上进行领域式搜索，由于该空间是一种面向地形本身分布特点和性质的地形分割，解路径的粒度较粗，因此又进行了在栅格规划空间上的第二次搜索──栅格搜索，以对解路径进行细化，最佳结点只在三个方向上扩展，并取代价最小者。This paper presents the design and implementation of cross-country path planning for an autonomous land vehicle based on a heuristic search method. This path planning consists of two searches by using heuristic evaluation functions. The first is a territory-formula search in territory-formula planning space which is a terrain separating based on inherent distributing quality and property of terrain. This searching path is rough. The second is a grid search in grid planning space for getting the minute sitepath. The best node is the smallest cost one expanded in three directions.

关 键 词：路径规划 栅格搜索 面向对象 机器人 自主机器人 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]