English

应用科学学报 ›› 1995, Vol. 13 ›› Issue (4): 439-446.

• 论文 • 上一篇    下一篇

机器人多传感器数据融合的两级结构及途径

汪晓钢, 钱文瀚   

  1. 上海交通大学
  • 收稿日期:1994-05-23 修回日期:1994-10-14 出版日期:1995-12-31 发布日期:1995-12-31

TWO-LAYER ARCHITECTURE AND STRATEGIES FOR ROBITIC MULTISENSOR DATA FUSION

WANG XIAOGANG, QTAN WENHAN   

  1. Shanghai Jiao Tong University
  • Received:1994-05-23 Revised:1994-10-14 Online:1995-12-31 Published:1995-12-31

PDF

88

被引次数

3

摘要： 机器人多传感器数据融合已成为一个活跃的研究领域,然而关于融合模型与途径的系统阐述是缺乏的。该文从信息处理的角度归纳出了一个数据信息融合的两级结构模型,数据级和证据级融合。这一模型较好反映了信息融合的本质.根据这一结构,又给出了运用Biyes参数估计和Dempster-Shafer证据推理相结合的多传感器数据融合的一般途径和步骤,最后加以例证。

关键词: 传感器融合, 估计理论, 不确定性推理, 信息融合

Abstract: In recent years,there has been increasing interest in the development of robotic distributed sensor networks,but little is known in the area of the architecture and strategies for multisensor data fusion.This paper constructs a model of two-layer architecture and strategies of sensor fusion of the data level and evidence level.The model embodies the principle of information processing in data fusion problems.According to the proposed structure,this paper describes the efficient algorithms for multisensor data combination and information fusion usin Bayes parameter estimation and Dempster-Shafer's theory.An example is considered in detail.

Key words: sensor fusion, information fusion, estimation theory, uncertainty reasoning