GLOBAL LOCALIZATION FOR MOBILE ROBOTS BY MULTIPLE HYPOTHESIS TRACKING

被引：10

作者：

PIASECKI, M

机构：

[1] Institute of Technical Cybernetics, Technical University of Wroclaw, 50-372 Wroclaw

来源：

ROBOTICS AND AUTONOMOUS SYSTEMS | 1995年 / 16卷 / 01期

关键词：

MOBILE ROBOT; LOCALIZATION; KALMAN FILTERING; MULTIPLE HYPOTHESIS TRACKING; GRAPH SEARCH;

D O I：

10.1016/0921-8890(95)00037-G

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

The paper presents a long-term localization method for autonomous mobile robot navigation in a known environment. The method uses observations from an ultrasonic range finder but is more global than standard incremental localization based on EKF. A probabilistic sensor fusion technique (Kalman filtering) is used to obtain local robot position update (local localization). Multiple hypothesis tracking is used to manage global localization, i.e. to determine the source of an observed signal or to point out an approximate robot position to initiate local position estimation. Results of computer simulation are presented.

引用

页码：93 / 104

页数：12