最优算术平均融合及其在非同视域场景的应用

doi:10.12305/j.issn.1001-506X.2025.06.02

摘要/Abstract

摘要：

为了在视域(field of view, FOV)不同的条件下实现对数量时变的不确定目标的最优跟踪, 提出一种高斯混合概率假设密度(Gaussian mixture probability hypothesis density, GM-PHD)滤波器的去相关算术平均(arithmetic average, AA)融合算法。鉴于多目标AA融合被分解为多组单目标分量的合并, 先通过重构贝叶斯融合推导出最优去相关估计融合, 后将其用作单目标分量的合并方法。由于推导的去相关估计融合需要先验估计, 设计了一个包含主滤波器的分层结构, 以自动提供需要的先验估计。为了解决不同FOV导致的目标势低估问题, 融合节点利用FOV补偿单目标分量的权重。仿真结果证实了提出的算法在各种场景中的最优性, 提升了多目标跟踪的精度。

关键词: 概率假设密度滤波器, 去相关, 贝叶斯融合, 分层结构, 主滤波器

Abstract:

A decorrelation arithmetic average (AA) fusion algorithm of Gaussian mixture probability hypothesis density (GM-PHD) filters is proposed to achieve optimal tracking of a time-varying number of uncertain targets within different field of view (FOV). Given that the multi-target AA fusion is decomposed into multiple groups of single-target component merging by association operation, optimal decorrelation estimation fusion is firstly derived by reshaping the Bayesian fusion and then is applied as the merging method of single-target components. Since the derived decorrelation estimation fusion requires prior estimates, a hierarchical structure involving a master filter dedicated to automatically providing prior estimates is designed. To address the underestimated target cardinality arising from different FOV, the fusion node compensates for weight of single-target components according to FOV. Simulation results demonstrate the proposed algorithm's optimality in various scenarios, which improves the multi-target tracking accuracy.

Key words: probability hypothesis density (PHD) filter, decorrelation, Bayesian fusion, hierarchical structure, master filter

中图分类号:

TP911

薛昱, 冯西安. 最优算术平均融合及其在非同视域场景的应用[J]. 系统工程与电子技术, 2025, 47(6): 1739-1745.

Yu XUE, Xi'an FENG. Optimal arithmetic average fusion and its application in different fields of view[J]. Systems Engineering and Electronics, 2025, 47(6): 1739-1745.

图/表 10

图1

表1

图2

图3

图4

图5

图6

表2

图7

图8

参考文献 30

1	闫莉萍, 刘晗钊, 夏元清. 天空地一体化多目标跟踪算法研究综述[J]. 信号处理, 2024, 40 (11): 1951- 1971. doi: 10.12466/xhcl.2024.11.003
	YAN L P , LIU H Z , XIA Y Q . Survey on space-air-ground integrated multi-target tracking algorithms[J]. Journal of Signal Processing, 2024, 40 (11): 1951- 1971. doi: 10.12466/xhcl.2024.11.003
2	王志伟, 刘永祥, 杨威, 等. 基于航迹概率假设密度的多传感器多目标跟踪[J]. 系统工程与电子技术, 2024, 46 (2): 526- 533. doi: 10.12305/j.issn.1001-506X.2024.02.17
	WANG Z W , LIU Y X , YANG W , et al. Multi-sensor multi-target tracking with trajectory probability hypothesis density[J]. Systems Engineering and Electronics, 2024, 46 (2): 526- 533. doi: 10.12305/j.issn.1001-506X.2024.02.17
3	MAHLER R . PHD filters of higher order in target number[J]. IEEE Trans.on Aerospace and Electronic Systems, 2007, 43 (4): 1523- 1543. doi: 10.1109/TAES.2007.4441756
4	VO B T , VO B N , CANTONI A . The cardinality balanced multi-target multi-Bernoulli filter and its implementations[J]. IEEE Trans.on Signal Processing, 2008, 57 (2): 409- 423.
5	JIANG Q , WANG R , NI N , et al. A Gaussian mixture PHD filter for multitarget tracking in target-dependent false alarms[J]. IEEE Trans.on Aerospace and Electronic Systems, 2024, 60 (4): 4808- 4824. doi: 10.1109/TAES.2024.3382068
6	SONMEZ H H , HOCAOGLU A K . Holder divergence-based reward function for Poisson RFSs and application to multitarget sensor management[J]. IEEE Sensors Journal, 2023, 23 (9): 9999- 10008. doi: 10.1109/JSEN.2023.3255987
7	UNEY M , CLARK D E , JULIER S J . Distributed fusion of PHD filters via exponential mixture densities[J]. IEEE Journal of Selected Topics in Signal Processing, 2013, 7 (3): 521- 531. doi: 10.1109/JSTSP.2013.2257162
8	BATTISTELLI G, CHISCI L, FANTACCI C, et al. Distributed fusion of multitarget densities and consensus PHD/CPHD filters[C]// Proc. of the Signal Processing Sensor/Information Fusion, and Target Recognition XXIV, 2015: 122-136.
9	LI G C , BATTISTELLI G , YI W , et al. Distributed multi-sensor multi-view fusion based on generalized covariance intersection[J]. Signal Processing, 2020, 166, 107246. doi: 10.1016/j.sigpro.2019.107246
10	JIN Y W , LI J X . Time-space domain assignment for generali-zed covariance intersection fusion with labeled multitarget densities[J]. Digital Signal Processing, 2023, 132, 103786. doi: 10.1016/j.dsp.2022.103786
11	UNEY M , HOUSSINEAU J , DELANDE E , et al. Fusion of finite-set distributions: Pointwise consistency and global cardinality[J]. IEEE Trans.on Aerospace and Electronic Systems, 2019, 55 (6): 2759- 2773. doi: 10.1109/TAES.2019.2893083
12	GAO L , BATTISTELLI G , CHISCI L . Multiobject fusion with minimum information loss[J]. IEEE Signal Processing Letters, 2020, 27, 201- 205. doi: 10.1109/LSP.2019.2963817
13	吴孙勇, 郑翔飞, 李天成, 等. 基于算术平均融合的分布式多伯努利扩展目标跟踪[J]. 电子与信息学报, 2023, 45 (6): 2171- 2179.
	WU S Y , ZHENG X F , LI T C , et al. Distributed multi-Bernoulli extended targets tracking based on arithmetic average fusion[J]. Journal of Electronics & Information Technology, 2023, 45 (6): 2171- 2179.
14	LI T C , YAN R B , DA K , et al. Arithmetic average density fusion-part Ⅲ: heterogeneous unlabeled and labeled RFS filter fusion[J]. IEEE Trans.on Aerospace and Electronic Systems, 2024, 60 (1): 1023- 1034. doi: 10.1109/TAES.2023.3334223
15	LI T C , LIANG H Z , LI G C , et al. Arithmetic average density fusion-Part Ⅳ: distributed heterogeneous fusion of RFS and LRFS filters via variational approximation[J]. IEEE Trans.on Signal Processing, 2024, 73, 1454- 1469.
16	LI T C , CORCHADO J M , PRIETO J . Convergence of distributed flooding and its application for distributed Bayesian filtering[J]. IEEE Trans.on Signal and Information Processing Over Networks, 2016, 3 (3): 580- 591.
17	DA K , LI T C , ZHU Y F , et al. A computationally efficient app- roach for distributed sensor localization and multitarget trac- king[J]. IEEE Communications Letters, 2019, 24 (2): 335- 338.
18	WEI J X , LUO F , QI J W , et al. Distributed fusion of labeled multi-Bernoulli filters based on arithmetic average[J]. IEEE Signal Processing Letters, 2024, 31, 656- 660. doi: 10.1109/LSP.2024.3364506
19	LI T C . Arithmetic average density fusion-part Ⅱ: unified derivation for unlabeled and labeled RFS fusion[J]. IEEE Trans. on Aerospace and Electronic Systems, 2024, 60 (3): 3255- 3268. doi: 10.1109/TAES.2024.3359592
20	YANG H , LI T C , YAN J K , et al. Hierarchical average fusion with GM-PHD filters against FDI and DoS attacks[J]. IEEE Signal Processing Letters, 2024, 31, 934- 938. doi: 10.1109/LSP.2024.3356823
21	WANG X D , SUN S D , LI T C , et al. Fault tolerant multi-robot cooperative localization based on covariance union[J]. IEEE Robotics and Automation Letters, 2021, 6 (4): 7799- 7806. doi: 10.1109/LRA.2021.3100000
22	LI T C , WANG X X , LIANG Y , et al. On arithmetic average fusion and its application for distributed multi-Bernoulli multitarget tracking[J]. IEEE Trans.on Signal Processing, 2020, 68, 2883- 2896.
23	LI T C , SONG Y , SONG E B , et al. Arithmetic average density fusion-part Ⅰ: some statistic and information-theoretic results[J]. Information Fusion, 2024, 104, 102199. doi: 10.1016/j.inffus.2023.102199
24	WANG K W , ZHANG Q , HU X L . Multi-sensor multi-target tracking arithmetic average fusion method based on probabili-stic time window[J]. IEEE Sensors Journal, 2024, 24 (3): 3583- 3593. doi: 10.1109/JSEN.2023.3337267
25	YI W , LI G C , BATTISTELLI G . Distributed multi-sensor fusion of PHD filters with different sensor fields of view[J]. IEEE Trans.on Signal Processing, 2020, 68, 5204- 5218. doi: 10.1109/TSP.2020.3021834
26	傅嘉政. 基于航迹随机集的非同视域稳健分布式融合算法[D]. 成都: 电子科技大学, 2023.
	FU J Z. Trajectory RFS based robust distributed fusion under different fields-of-view[D]. Chengdu: University of Electronic Science and Technology of China, 2023.
27	CHEN X L , ZHANG L Y , GU C Q , et al. Wideband Sherman-Morrison-Woodbury formula-based algorithm for electromagnetic scattering problems[J]. IEEE Trans.on Antennas and Propagation, 2023, 71 (6): 5487- 5492. doi: 10.1109/TAP.2023.3263627
28	LI T C , CORCHADO J M , SUN S D . Partial consensus and conservative fusion of Gaussian mixtures for distributed PHD fusion[J]. IEEE Trans.on Aerospace and Electronic Systems, 2018, 55 (5): 2150- 2163.
29	ZHANG H W , LI J , ZHANG J R , et al. Speeding up k-means clustering in high dimensions by pruning unnecessary distance computations[J]. Knowledge-Based Systems, 2024, 284, 111262. doi: 10.1016/j.knosys.2023.111262
30	SCHUHMACHER D , VO B T , VO B N . A consistent metric for performance evaluation of multi-object filters[J]. IEEE Trans.on Signal Processing, 2008, 56 (8): 3447- 3457. doi: 10.1109/TSP.2008.920469

编号i	初始状态x₁ⁱ/(m, m, m/s, m/s)	存活时间/s
目标-1	700, 950, 2.5, -8	2~100
目标-2	300, 700, 12, -2	13~80
目标-3	300, 700, 25, 5	25~60
目标-4	1 200, 900, 3, -16	30~55
目标-5	1 200, 400, 4, 30	35~50
目标-6	800, 300, 1, 10	10~70

算法	均值	最大值	标准差
GM-PHD	6.684 9	9.334 6	1.496 9
标准AA	6.210 6	9.306 9	1.177 7
补偿权重GA	5.189 6	9.046 0	1.235 3
补偿权重-AA	4.961 5	8.516 1	1.291 6
去相关AA	4.273 2	8.326 5	1.175 6

[1]	王森, 鲍庆龙, 潘嘉蒙, 祝茜. 基于改进概率假设密度滤波器的非合作双基地雷达目标跟踪[J]. 系统工程与电子技术, 2023, 45(7): 2002-2009.
[2]	邓嘉宁, 吴宇, 许舒婷, 苟进展. 基于模糊贝叶斯-ANP舰载机出动回收综合评估[J]. 系统工程与电子技术, 2022, 44(11): 3423-3432.
[3]	焦志强, 张杰勇, 姚佩阳, 王勋, 何宜超. 基于分层结构的C4ISR服务部署分布式演化方法[J]. 系统工程与电子技术, 2021, 43(6): 1572-1585.
[4]	甄岩, 李兴, 王汝言. 带有群体智能的能量感知分层数据转发机制[J]. 系统工程与电子技术, 2017, 39(9): 2109-2118.
[5]	朱友清, 周石琳, 高贵. 结合聚类的GM-PHD滤波器辐射源群目标跟踪[J]. 系统工程与电子技术, 2015, 37(9): 1967-1973.
[6]	朱友清, 周石琳. 基于雷达辐射源信号特征的类别信息辅助GM-PHD滤波器[J]. 系统工程与电子技术, 2015, 37(6): 1273-1279.
[7]	杨峰，王永齐，梁彦，潘泉. 面向快速多目标跟踪的协同PHD滤波器[J]. 系统工程与电子技术, 2014, 36(11): 2113-2121.
[8]	连峰，元向辉，陈辉. 基于势概率假设密度滤波器的不可分辨目标跟踪算法[J]. 系统工程与电子技术, 2013, 35(12): 2445-2451.