一种室内自主移动机器人定位方法  

Localization method in indoor environment for autonomous mobile robot

作　　者：高云峰[1] 吕明睿 周伦[1] 霍光磊[1] 

机构地区：[1]哈尔滨工业大学机电工程学院,黑龙江哈尔滨50001

出　　处：《华中科技大学学报（自然科学版）》2013年第S1期245-248,253,共5页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：机器人技术与系统国家重点实验室自主课题(SKLRS201201B);国家科技支撑计划资助项目(2012BAI33B04)

摘　　要：针对常规超声波传感器网络传感器节点布置密度大的不足,提出了一种基于单对超声波传感器的非同步到达时间(ATOA)测距的定位方法.该方法利用了分时到达的超声波测距信息和航迹推算得到的位移矢量信息获得自主移动机器人的起始位姿信息.分析了该方法中各传感器的误差因素对计算结果的影响,并根据分析结果提出了一种圆周定位的优化方案.同时还提出了一种双层卡尔曼滤波器来融合多组定位测算结果的算法,提高了机器人定位精度,实现了机器人轨迹跟踪.计算机仿真结果表明使用圆周定位方法和双层卡尔曼滤波算法能有效地在较短时间内获得自主移动机器人较高的位姿精度.Aiming at overcoming the drawback of ordinary ultrasonic sensor network in which a high density of sensor nodes are required,a localization method using asynchronous-time-of-arrival(ATOA)measurements from single ultrasonic sensor pair was proposed.This method utilized the ultrasonic distance measurements obtained sequentially and displacement vector measurements from dead reckoning(DR)to calculate the launching pose of autonomous mobile robots(AMRs).Error factors that influenced the result of ATOA were investigated and an optimizing scheme of circle localization was proposed based on the analysis of errors.In order to fuse couples of calculating results of ATOA, a double-layer Kalman filter(DLKF)was developed to track robot and enhance positioning precision. The computer simulation was conducted to prove that AMRs can obtain precise pose information in a short time by using circle localization and the DLKF.

关 键 词：超声波定位 航迹推算 双层卡尔曼滤波 信息融合 多传感器技术 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]