摘要
本文在传统粒子滤波的基础上提出了一种基于自适应感知复位定位算法(ASRL:Adaptive Sensor Reset- ting Localization)的移动机器人定位方法.该方法通过带有权值的样本集描绘机器人的可信度,根据有效的样本数计算需要生成的新样本数,然后从感知分布中采样代替原来的样本.ASRL算法已经在安装有编码器和彩色摄像机两种传感器的实际移动机器人AMR-ITL上进行实验,结果表明该算法鲁棒性更好,收敛更快.
An adaptive Sensor Resetting Localization(ASRL)algorithm based on traditional Particle Filter is proposed for mobile robot.The belief of robot is represented by a set of weighted samples,new necessary samples are calculated according to effective samples and resampled based on sensor data,and then old samples are replaced with new samples during ASRL algorithm. This algorithm is used on autonomous mobile robot AMR-ITL equipped with encoder and color camera sensor successfully,experiment result shows that ASRL is a more robust and quickly convergence algorithm.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2007年第11期2166-2171,共6页
Acta Electronica Sinica
基金
中国民航大学人才启动基金(No.QD02X15)
关键词
移动机器人
定位
粒子滤波
自适应感知复位
mobile robot
localization
particle filter
adaptive sensor resetting localization