摘要
作为智能交通的重要组成部分,智能车辆旨在辅助或取代驾驶员进行车辆驾驶,以减少交通事故,提高交通系统效率.本文针对半结构化道路下基于视觉导航智能车辆的路径规划和车辆控制问题进行了研究试验.首先采用局部逆透视变换方法将存在无穷视野的智能车辆成像系统转换为真实世界坐标表示;然后利用深度优先搜索算法搜索当前道路环境下所有可行路径,依据无碰撞约束条件设定行驶轨迹;以预设轨迹的行驶时间和视觉图像信息作为输入,设计分级模糊控制器进行车辆横向控制;最后给出场地试验结果.试验表明,本文路径规划和车辆控制算法有效稳定,实现了校园环境下障碍物躲避、路径规划和车辆自主行驶的设计功能.
As a pivotal component of intelligent transportation system,intelligent vehicles aim to reduce traffic accidents and improve transportation system efficiency,eventually,accomplish the objective of assisting or totally replacing drivers to control vehicles.In this paper,a research is conducted on path planning and vehicle control for vision-based navigation intelligent vehicles under semi-structured road condition.Firstly,local inverse perspective transformation is introduced to transfer vision imaging system of intelligent vehicle,which contains infinite vision field,into real world coordinates system;then depth-first search method is used to calculate all accessible paths in current road environment.The travel route,meanwhile,is preset under constraint of collision-free;after that,the travel time and vision image information are taken as inputs of a hierarchical fuzzy controller which is designed for vehicle control;finally,field test in real-road environment demonstrates the system not only satisfies all the design functions,including obstacle avoiding,path planning and vehicle autonomous driving,but also achieves a high effectiveness and stability.
出处
《交通运输系统工程与信息》
EI
CSCD
2011年第2期44-51,共8页
Journal of Transportation Systems Engineering and Information Technology
关键词
智能交通
智能车辆
半结构化道路
逆透视变换
路径规划
模糊控制
intelligent transportation
intelligent vehicles
semi-structured road
inverse perspective transformation
path planning
fuzzy control
