改进路标定位传感模型在机器人定位中的应用  

Improved Localizition Sensor Model Based on Landmarks and Its Application in Robot

作　　者：申伟锋[1] 钱东海[1] 胡猛[1] 

机构地区：[1]上海大学精密机械系,上海200072

出　　处：《机械设计与制造》2013年第8期74-77,共4页Machinery Design & Manufacture

摘　　要：分析了移动机器人基于路标定位的贝叶斯滤波理论中的传感模型,传统传感模型利用周期末检测到的观测值进行更新修正。若在一个周期内不同位置检测到多个不同角度观测值,传统的传感模型默认为这些角度观测值为周期末静止时所得到,提出了通过转化算法将这些角度观测值转换到更新周期末,使得周期末角度观测值得到的位置观测更精确。计算机实验数据表明提出的传感模型观测值转换算法与现有传统算法传感模型相比,更新位置具有更高的精度。It analyzes the sensor model of baysian filtering about robot localization based on landmarks,and the tradition sensor model used observed values for updating.Different angle observed values were detected in different location when laser scanned a cycle.The traditional sensor model defaulted that these angle observed values were detected in update point（an cycle end）.It proposed that these observed values were transformed to the cycle end by algorithm,and the algorithm made higher observed location accuracy.The experiment carried out by computer shows that the algorithm proposed in this paper has higher localization accuracy than the traditional sensor model algorithms.

关 键 词：机器人 路标定位 贝叶斯理论 传感模型 观测值转换 

分 类 号：TH16[机械工程—机械制造及自动化] TP24[自动化与计算机技术—检测技术与自动化装置]