兵器工业

 基于MIMOFNN模型的弹道导弹目标
时空序贯融合识别方法

展开
  •  1.空军预警学院, 武汉 430019; 2. 中国人民解放军第66132部队, 北京 100043;
    3. 福建农林大学 东方学院计算机系, 福州 350017

网络出版日期: 2017-09-20

基金资助

 国家自然科学基金青年科学基金(61401503)资助

 TemporalSpatial Sequential Fusion Recognition Method of
 Ballistic Missile Target Based on MIMOFNN Model

Expand
  •  1. Air Force Early Warning Academy, Wuhan 430019, China;
    2. Unit No.66132 of PLA, Beijing 100043, China; 3. Department of Computer Science,
     Dongfang College, Fujian Agriculture and Forestry University, Fuzhou 350017, China

Online published: 2017-09-20

Supported by

 

摘要

 针对传统弹道导弹(BM)目标时空序贯融合识别算法识别效率低、抗噪性能差的缺点,提出了一种基于多输入多输出模糊神经网络(MIMOFNN)模型的BM目标时空序贯融合识别方法.该方法首先根据多层融合的思想,结合神经网络和模糊理论,提出了多传感器多特征MIMOFNN模型;其次,在此基础上,将当前时刻的融合结果与下一时刻的融合结果再融合,得到此时刻时空序贯融合识别结果,并将其与识别门限比较,直到满足识别门限要求,时空序贯融合识别结束,并做出决策;最后通过实验验证了所提模型的有效性和良好的抗噪性.

本文引用格式

李昌玺1,2,周焰1,林菡3,李灵芝1,郭戈1 .  基于MIMOFNN模型的弹道导弹目标
时空序贯融合识别方法[J]. 上海交通大学学报, 2017
, 51(9) : 1138 . DOI: 10.16183/j.cnki.jsjtu.2017.09.018

Abstract

 In traditional temporalspatial sequential fusion recognition method of ballistic missile (BM) target, there always exits the problem that the efficiency is low and the antinoise performance is bad. In order to solve these difficulties, this paper proposed a fusion recognition method which is called temporalspatial sequential fusion recognition method of BM target based on multiple input multiple output and fuzzy neural network (MIMOFNN) model. In this model, firstly, we use the idea of multi layer fusion, combine with neural network and fuzzy theory, and put forward the MIMOFNN model with multi sensors of multi features. And then, we reintegrate the results of present moment and next moment to get the fusion results. Meanwhile, we compare with recognition threshold, until the fusion results match the recognition threshold, and the process of fusion ends. Finally, the experiment validates the effectiveness and antinoise performance of this model.

参考文献

 [1]周万幸. 弹道导弹雷达目标识别技术[M]. 北京:电子工业出版社,2011.
[2]吴瑕, 周焰, 杨龙坡,等. 基于聚类分析与几何的目标特征敏感性评估算法[J]. 控制与决策, 2012, 27(6): 914918.
WU Xia, ZHOU Yan, YANG Longpo, et al. Target feature sensitivity evaluation method based on clustering analysis and geometry[J]. Control and Decision, 2012, 27(6): 914918.
[3]吴瑕,周焰,蔡益朝,等.多传感器目标融合识别系统模型研究现状与问题[J].字航学报, 2010, 31(5): 14131420.
WU Xia, ZHOU Yan, CAI Yichao, et al. Research actualities and problems on multisensor target recognition system model[J]. Journal of Astronautics,2010, 31(5): 14131420.
[4]焦朋勃, 黄长强, 蔡佳, 等. 基于时空信息融合的复合引信目标识别方法[J]. 探测与控制学报, 2014, 36(5): 4750.
JIAO Pengbo, HUANG Changqiang, CAI Jia, et al. Target recognition of compound fuse based on temporalspatial information fusion[J]. Journal of Detection & Control, 2014, 36(5): 4750.
[5]洪昭艺, 高勋章, 黎湘. 基于DS理论的混合式时空域信息融合模型[J].信号处理, 2011, 27(1): 1419.
HONG Zhaoyi, GAO Xunzhang, LI Xiang. Research on temporalspatial information fusion model based on DS theory[J]. Signal Processing, 2011, 27(1): 1419.
[6]王宇晨, 胡晓伟, 李伟龙. 基于时空序贯融合的TBM拦截效果评估[J]. 火力与指挥控制, 2015, 40(2): 100104.
WANG Yuchen, HU Xiaowei, LI Weilong. Intercepting effect evaluation of TBM based on temporalspatial information fusion[J]. Fire Control & Command Control, 2015, 40(2): 100104.
[7]吴俊, 程咏梅, 曲圣杰, 等. 基于三级信息融合结构的多平台多雷达目标识别算法[J]. 西北工业大学学报, 2012, 30(3): 367372.
WU Jun, CHENG Yongmei, QU Shengjie, et al. An effective multiplatform multiradar target identification algorithm based on three level fusion hierarchical structures[J]. Journal of Northwestern Polytechnical University, 2012, 30(3): 367372.
[8]朱玉鹏, 付耀文, 黎湘,等. 决策层时空信息融合的神经网络模型研究[J].系统工程与电子技术, 2008, 30(6): 10981102.
ZHU Yupeng, FU Yaowen, LI Xiang, et al. Research on a new network model for temporalspatial information fusion at decision level[J]. Systems Engineering and Electronics, 2008, 30(6): 10981102.
[9]罗大庸, 张远. 多传感器信息时空融合模型及算法研究[J].系统工程与电子技术, 2004, 26(1): 3639.
LUO Dayong, ZHANG Yuan. Research of spatialtemporal architecture model and the algorithm for multisensor information fusion[J]. Systems Engineering and Electronics, 2004, 26(1): 3639.
Options
文章导航

/