摘要
动态环境下运动物体的检测是移动机器人研究的难点问题之一。以未知环境为研究背景,提出了一种基于激光雷达的自主动态障碍检测方法。通过k-近邻方法对激光雷达测距数据进行了空间障碍聚类,在此基础上分析了聚类障碍的特征参数,例如区域、质心,利用聚类障碍数据的时间关联性分析并确定了障碍的类型,并采用模糊地图匹配的策略实现了地图匹配和更新。在所研制的移动机器人上进行了实验,实验结果验证了方法的有效性。
Detection of moving objects under dynamic environments is a difficult problem in mobile robot research.Considering the unknown environment as research background,an autonomous detection method of dynamic obstacles based on laser scanner is presented.The k-nearest neighbor approach is adopted to realize the spatial clustering of obstacles according to the ranging data of laser scanner.The characteristic parameters of clustering obstacles such as area and centroid are analyzed.With the analysis of time association,the types of obstacles are determinated.In order to update the environmental map,the map matching strategy using fuzzy logic is utilized.Experiments with the mobile robot are implemented.The experimental result verifies the validity of this approach.
出处
《光学技术》
EI
CAS
CSCD
北大核心
2008年第2期289-293,共5页
Optical Technique
关键词
激光雷达
移动机器人
动态障碍检测
laser scanner
mobile robot
detection of dynamic obstacles