室外无人车自主导航方法研究

发布时间：2018-01-11 23:15

本文关键词：室外无人车自主导航方法研究　出处：《浙江大学》2017年硕士论文　论文类型：学位论文

【摘要】：本文介绍了室外环境中无人车的自动导航方法研究和系统设计工作。针对室外未知环境,依托车载的双目相机和激光雷达,对环境进行安全的自主探索和感知建模,并重用环境模型信息进行自主定位和规划控制,完成安全的点到点自主导航。本文的特色是:主要基于立体视觉,即实现了环境建模、自定位、自主避障等自主导航系统的核心功能,并基于稀疏视觉地图的定位信息作为连接各个导航环节的纽带,完成整个导航系统的构建。主要研究成果包括:1.提出了从稀疏视觉地图构建度量-拓扑混合地图的方法。2.提出了基于上述地图的无人车位置跟踪和实时避障算法。3.开发了一套小型无人车软硬件系统,集成了环境感知、规划控制、自主避障等功能。
[Abstract]:This paper introduces the research of the automatic navigation method and the system design of the unmanned vehicle in the outdoor environment. Aiming at the unknown outdoor environment, it relies on the binocular camera and the lidar in the vehicle. The environment is independently explored and perceived, and the environment model information is reused for autonomous positioning and planning control to complete the safe point-to-point autonomous navigation. The main features of this paper are: based on stereo vision. That is to achieve the environmental modeling, self-positioning, autonomous obstacle avoidance and other autonomous navigation system core functions, and based on sparse visual map location information as a link between the various navigation links. Complete the construction of the whole navigation system. The main research results include:. 1. The method of constructing metric topological hybrid map from sparse visual map is presented. 2. A position tracking and real-time obstacle avoidance algorithm based on above map is proposed. 3. A software and hardware system of small unmanned vehicle is developed. Tong. Integrated environmental awareness, planning control, autonomous obstacle avoidance and other functions.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN96

【相似文献】