欢迎访问《空军工程大学学报》官方网站!

咨询热线:029-84786242 RSS EMAIL-ALERT
“集中式融合的性能一定优于分布式融合的性能”存疑
DOI:
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

TN953

基金项目:

国家自然科学基金资助项目(60774091);陕西省自然科学基础研究计划资助项目(2007F24)


Doubt abut Assertion of ‘Performance of Centralized Fusion Must Be Better Than That of Distributed Fusion’
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    信息融合领域存在一个公认的“定论”,即集中式融合的性能一定比分布式融合的好,且是最优的;业已证明,当融合系统采用单模型滤波算法,如卡尔曼滤波器时,上述“定论”是成立的;那么在融合系统采用多模型滤波算法,如针对机动目标跟踪的交互多模型滤波(IMM)算法时,情况是否也是如此?抱着对此存疑的态度,对采用IMM算法时的集中式融合与分布式融合进行了研究,给出了扩维、序贯以及等效量测3种集中式融合算法和简单方差凸组合、互协方差组合2种分布式融合算法,并设置不同运动场景对算法进行了大量仿真。仿真实验结果表明,针对交互多模型滤波估计的集中式融合性能并不一定比采用同样滤波算法的分布式融合的好,并给出了作者对此“诧异”现象的认识和理解。

    Abstract:

    In information field, one famous assertion is that performance of centralized fusion must be better than that of distributed fusion and also is optimal. It has been proven that this assertion is correct as single model filter, such as kalman filter, is adopted in fusion system. However, in case of maneuvering target tracking, in which multiple model filter, such as IMM filter, is usually applied in fusion system, is above assertion still valid? With doubtful attitude, this problem is investigated in the paper, three centralized fusion algorithms and two distributed fusion algorithms when IMM is used are presented, and simulations are made in different kinematic sceneries. Simulation results indicate that, as IMM filter is used in fusion system, performances of centralized fusion are not always better than those of distributed fusion. Analyses about this surprise result are also given in last section of the paper.

    参考文献
    相似文献
    引证文献
引用本文

乔向东,李涛,张志伟,杨仝,李鸿艳,冯新喜.“集中式融合的性能一定优于分布式融合的性能”存疑[J].空军工程大学学报,2010,(5):53-59

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2015-11-17
  • 出版日期: