系统工程与电子技术 ›› 2023, Vol. 45 ›› Issue (4): 1082-1089.doi: 10.12305/j.issn.1001-506X.2023.04.16

• 系统工程 • 上一篇    

海上动目标身份置信度融合计算模型

任文娟1,2,*, 杨战鹏1,2, 许光銮1,2, 付琨1,2   

  1. 1. 中国科学院空天信息创新研究院, 北京 100094
    2. 中国科学院网络信息体系技术重点实验室, 北京 100190
  • 收稿日期:2022-04-28 出版日期:2023-03-29 发布日期:2023-03-28
  • 通讯作者: 任文娟
  • 作者简介:任文娟 (1982—), 女, 副研究员, 博士, 主要研究方向为多源信息融合、雷达辐射源识别
    杨战鹏 (1988—), 男, 助理研究员, 博士, 主要研究方向为泛函分析、微分方程、迁移学习
    许光銮 (1978—), 男, 研究员, 博士, 主要研究方向为地理空间大数据挖掘、遥感图像理解
    付琨 (1974—), 男, 研究员, 博士, 主要研究方向为地理空间大数据挖掘、遥感图像理解

Fusion calculation model of sea moving target identity confidence

Wenjuan REN1,2,*, Zhanpeng YANG1,2, Guangluan XU1,2, Kun FU1,2   

  1. 1. Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China
    2. Key Laboratory of Network Information System Technology, Chinese Academy of Sciences, Beijing 100190, China
  • Received:2022-04-28 Online:2023-03-29 Published:2023-03-28
  • Contact: Wenjuan REN

摘要:

为了支持海上动目标实时打击决策, 提出基于D-S(Dempster-Shafer)证据理论的移动目标身份置信度动态演化与融合框架, 构建海上动目标身份置信度多源融合计算模型, 实现目标身份置信度连续可靠评估。针对移动目标的多源异步信息, 建立目标身份辨识空间, 构建历史与当前目标信息的基本概率分配函数, 表征历史信息作为证据随时间推演而不断增加的不确定性, 利用历史与当前二元证据合成规则计算目标身份融合置信度。仿真结果表明, 所提方法能够提高目标身份置信度、降低信息融合系统目标识别的不确定性, 扩展了目标打击窗口, 提升了决策可靠性和打击灵活性。

关键词: 信息融合, 多源异步信息, 移动目标, D-S证据理论, 身份置信度

Abstract:

In order to support the real-time attack decision of moving targets at sea, dynamic evolution and fusion framework of moving target identity confidence based on D-S (Dempster-Shafer) evidence theory is proposed. A multi-source fusion calculation model of sea moving target identity confidence is constructed to realize the continuous and reliable evaluation of target identity confidence. To cope with the multi-source asynchronous information of moving targets, the target identification space is established, in which the basic probability distribution function of historical and current target information is constructed and used to characterize the increasing uncertainty of historical information as evidence deduced over time. In addition, the target identity fusion confidence is calculated by using the historical and current binary evidence synthesis rules. Simulation results show that the proposed method can improve the target identity confidence, reduce the uncertainty of target recognition in information fusion system, expand the target attack window, and improve the reliability of decision-making and the flexibility of attack.

Key words: information fusion, multi-source asynchronous information, moving target, Dempster-Shafer (D-S) evidence theory, identity confidence

中图分类号: