一种新的面向无人车的层级智能导航地图架构
首发时间:2017-05-16
摘要:针对无人车辆在半结构、非结构化局部任务区域的智能地图导航问题,提出一种新的层级智能导航地图架构,架构按照导航智能的层级分为三层,最底层为基础层,用于构建道路环境的基础信息知识库;中间层为导引层,根据知识库信息及车载定位系统为无人车其他控制模块提供导引信息;最上层为学习层,利用机器学习方法,学习人的驾驶经验,使无人车在面临一些已知场景时能够运用所学的经验规则进行决策。实验与实际测试表明,该架构能够较好满足无人车局部任务区域智能导航的要求,具有较好的扩展性和实际应用价值。
For information in English, please click here
A New Architecture of Level Intelligent Navigation Map for Unmanned Vehicles
Abstract: Pertinent to the intelligent mapping navigation of unmanned vehicles in semi-structural and non-structural partial task area, the paper proposes a new architecture of level intelligent navigation map, which is classified into three levels according to the hierarchies of intelligent navigation. The bottom is the base layer, which provides instructions for other control modules of unmanned vehicles according to the information base and vehicle positoning system. The upper is the learning layer, which absorbs humans' driving experience via robots so that unmanned vehicles can apply the experience and rules in decision-making when facing some known scenes. Experiments and actual tests indicate that the architecture can well meet unmanned vehicles's requirements on intelligent navigation in partial task areas and that it has satisfactory expansibility and actual application values.
Keywords: unmanned vehicle construction of navigation map robot learning information
论文图表:
引用
No.4732112119947714****
同行评议
勘误表
一种新的面向无人车的层级智能导航地图架构
评论
全部评论0/1000