基于机动频率自适应的增强层级融合算法

doi:10.12305/j.issn.1001-506X.2025.10.04

摘要/Abstract

摘要：

针对多源航迹长时间融合跟踪中因依赖经验参数而导致的滤波发散与航迹失真问题，提出一种基于机动频率自适应参数的增强层级融合算法。该算法改善传统融合方案对经验参数和理想接收环境的依赖，实现在复杂场景对目标的稳定跟踪并生成综合航迹。实验结果表明，结合机动频率自适应后的增强层级融合模型平均优化幅度达到36.16%。该算法通过引入机动频率驱动的自适应机制与层级反馈结构，显著增强了系统在高动态环境中的鲁棒性与时效性，可为复杂场景下多源航迹的高精度、稳定融合提供有效解决方案。

关键词: 机动频率自适应, 多传感器协同, 异构信息源联合, 增强层级融合模型

Abstract:

To address the issues of filter divergence and track distortion in long-term multi-source track fusion tracking caused by reliance on empirical parameters, an enhanced hierarchical fusion algorithm based on maneuver frequency adaptive parameters is proposed. This algorithm improves the dependence of traditional fusion methods on empirical parameters and ideal reception conditions, enabling stable target tracking and generation of integrated track in complex scenarios. Experimental results show that the enhanced hierarchical fusion model incorporating maneuver frequency adaptation achieves an average optimization improvement of 36.16%. This algorithm significantly enhances the robustness and timeliness of the system in high dynamic environments by introducing an adaptive mechanism driven by maneuvering frequency and a hierarchical feedback structure, providing an effective solution for high-precision and stable fusion of multi-source track in complex scenarios.

Key words: adaptive maneuvering frequency, multi sensor cooperation, heterogeneous information source fusion, enhanced hierarchical fusion model

中图分类号:

钟告知, 徐弘毅, 侯长波, 赵鹏旗, 郭浩南. 基于机动频率自适应的增强层级融合算法[J]. 系统工程与电子技术, 2025, 47(10): 3155-3167.

Gaozhi ZHONG, Hongyi XU, Changbo HOU, Pengqi ZHAO, Haonan GUO. Enhanced hierarchical fusion algorithm based on adaptive maneuvering frequency[J]. Systems Engineering and Electronics, 2025, 47(10): 3155-3167.

图/表 22

图1

图2

表1

表2

表3

表4

表5

图3

图4

图5

图6

表6

表7

图7

图8

表8

图9

表9

图10

图11

图12

图13

参考文献 31

1	KHANPARA M B, PATEL P B, PARMAR B N, et al. Transforming agriculture with drones: applications, challenges and implementation strategies[J]. Journal of Scientific Research and Reports, 2024, 30 (8): 792- 802. doi: 10.9734/jsrr/2024/v30i82299
2	FANG Z, SAVKIN V A. Strategies for optimized UAV surveillance in various tasks and scenarios: a review[J]. Drones, 2024, 8 (5): 193. doi: 10.3390/drones8050193
3	KUMAR A C, GIRIDHAR O. UAV detection multi-sensor data fusion[J]. Journal of Research in Science and Engineering, 2024, 6 (7): 6- 12. doi: 10.53469/jrse.2024.06(07).02
4	CHENG L, CHAO W, YAN P. IMCN: improved modular co-attention networks for visual question answering[J]. Applied Intelligence, 2024, 54 (6): 5167- 5182. doi: 10.1007/s10489-024-05456-4
5	MAHENDRA M, CHU K C, SANJEEV A, et al. Heterogeneous track-to-track fusion in 3-D using IRST sensor and air MTI radar[J]. IEEE Trans. on Aerospace and Electronic Systems, 2019, 55 (6): 3062- 3079. doi: 10.1109/TAES.2019.2898302
6	赵昊宸. 分布式雷达红外多传感器系统中的轨迹融合技术研究[D]. 长沙: 国防科技大学, 2017.
	ZHAO H C. Research on trajectory fusion technology in distributed radar/infrared multi sensor systems[D]. Changsha: National University of Defense Technology, 2017.
7	吴静, 艾小锋, 赵锋, 等. 航迹融合中的量测协方差统一转换方法[J]. 信息对抗技术, 2023, 2 (3): 74- 84.
	WU J, AI X F, ZHAO F, et al. Measurement covariance unified transformation method in trajectory fusion[J]. Information Adversarial Technology, 2023, 2 (3): 74- 84.
8	XIE Z L, ZHANG Z P, CHEN J L, et al. Data-driven unsupervised anomaly detection of manufacturing processes with multi-scale prototype augmentation and multi-sensor data[J]. Journal of Manufacturing Systems, 2024, 77 (3): 26- 39.
9	WANG X, WU Z X, JIN B, et al. MDSTF: a multi-dimensional spatio-temporal feature fusion trajectory prediction model for autonomous driving[J]. Complex & Intelligent Systems, 2024, 10 (5): 7419- 7420.
10	SPADON G, KUMAR J, EDEN D, et al. Multi-path long-term vessel trajectories forecasting with probabilistic feature fusion for problem shifting[J]. Ocean Engineering, 2024, 312 (15): 138.
11	XU Z Q, GUO Y F, KUANG Y, et al. Multi-sensor distributed fusion based on cross-location for passive tracking[J]. Signal, Image and Video Processing, 2024, 18, 9441- 9449.
12	LEE S, YUAN Z, PETRUNIN I, et al. Impact analysis of time synchronization error in airborne target tracking using a heterogeneous sensor network[J]. Drones, 2024, 8 (5): 167- 168. doi: 10.3390/drones8050167
13	YI P Z, TAO W, SHI Q Z, et al. Adaptive multi-pedestrian tracking by multi-sensor: track-to-track fusion using monocular 3D detection and MMW radar[J]. Remote Sensing, 2022, 14 (8): 1837. doi: 10.3390/rs14081837
14	ZAREI M, CHHABRA R, MOTTAGHI M. Sequential track fusion in multi-sensor networks of unscented Kalman filters: a case of slip estimation in planetary mobile robots[J]. Acta Astronautica, 2024, 219 (2): 555- 568.
15	SHENG X L, WANG Y, MU M F, et al. A modified asynchronous sequential Kalman track fusion based on sound propagation time[J]. EURASIP Journal on Advances in Signal Processing, 2023, 3 (1): 14- 13.
16	WANG M C, LI Y, MIN L, et al. Intelligent marine area supervision based on AIS and radar fusion[J]. Ocean Engineering, 2023, 285 (2): 5- 13.
17	XIE Z L, ZHANG Z P, CHEN J L, et al. Data-driven unsupervised anomaly detection of manufacturing processes with multi-scale prototype augmentation and multi-sensor data[J]. Journal of Manufacturing Systems, 2024, 77 (4): 26- 39.
18	邵俊伟. 一种参数自适应调整的Singer模型跟踪算法[J]. 科技信息, 2014, 15 (13): 111- 112.
	SHAO J W. A singer model tracking algorithm with parameter adaptive adjustment[J]. Science and Technology Information, 2014, 15 (13): 111- 112.
19	KASHANI B H. Deep attentive adaptive filter module in residual blocks for text-independent speaker verification[J]. Engineering Applications of Artificial Intelligence, 2024, 138 (A): 109312.
20	任少伟, 王睿, 张平定. 基于机动频率自适应的目标跟踪算法[J]. 空军工程大学学报（自然科学版）, 2004, (5): 32- 35.
	REN S W, WANG R, ZHANG P D. Target tracking algorithm based on adaptive maneuvering frequency[J]. Journal of Air Force Engineering University （Natural Science Edition）, 2004, (5): 32- 35.
21	WU Z, NI J G. Diffusion random Fourier adaptive filtering algorithm based on logistic distance metric for distributed estimation[J]. Digital Signal Processing, 2024, 156 (A): 104768.
22	PADMAPRIYA R, JEYASEKAR A. CA-EBM3D-NET: a convolutional neural network combined framework for denoising with weighted alpha parameter and adaptive filtering[J]. International Journal of Information Technology, 2024, 16, 4855- 4867.
23	ARASARATNAM I, HAYKIN S. Cubature Kalman filters[J]. IEEE Trans. on Automatic Control, 2009, 54 (6): 1254- 1269. doi: 10.1109/TAC.2009.2019800
24	WU W, JIANG J, LIU W J, et al. A sequential converted measurement Kalman filter in the ECEF coordinate system for airborne Doppler radar[J]. Aerospace Science and Technology, 2016, 51 (12): 11- 17.
25	XU L, MA P, SU X. Adaptive track association in distributed multi-sensor track fusion system[J]. ICIC Express Letters, 2011, 5 (11): 4101- 4106.
26	HYUN D, YOON H. A fusion algorithm considering error characteristics of the multi-sensor[J]. Journal of KIISE: Computer Systems and Theory, 2009, 36 (4): 274- 282.
27	WANG X, GUO R, JHA K N, et al. Multi-sensor targets data association and track fusion based on novel AWFCM[J]. International Journal of Nonlinear Sciences and Numerical Simulation, 2009, 10 (6): 811- 822.
28	SANKETH A, KUMAR S K, VPS N, et al. Airborne multi target track to track fusion of radar and IRST for advanced multi role combat aircrafts[J]. IFAC PapersOnLine, 2022, 55 (22): 224- 229. doi: 10.1016/j.ifacol.2023.03.038
29	WANG T L, ZHANG X A, YANG X L. A multiresolution adaptive grid method for solving the generalized probability density equation in stochastic dynamic analysis[J]. Acta Mechanica, 2022, 233 (5): 1911- 1940. doi: 10.1007/s00707-022-03183-w
30	YAN H, ZHANG G T, CUI W, et al. Multi-modality hierarchical fusion network for lumbar spine segmentation with magnetic resonance images[J]. Control Theory and Technology, 2024, 22 (4): 612- 622.
31	孔尚萍, 张海瑞, 廖选平, 等. 基于IMM-UKF的分层航迹融合算法[J]. 火力与指挥控制, 2018, 43 (2): 120- 124. doi: 10.3969/j.issn.1002-0640.2018.02.025
	KONG S P, ZHANG H R, LIAO X P, et al. Hierarchical trajectory fusion algorithm based on IMM-UKF[J]. Fire and Command Control, 2018, 43 (2): 120- 124. doi: 10.3969/j.issn.1002-0640.2018.02.025

变量	符号	变量描述	注释
输入变量	$ \left[ {L,B,H} \right] $	经度、纬度、高度列表	基于WGS-84坐标系
输出变量	$ \left[ {{X_c},{Y_c},{Z_c}} \right] $	X 轴、Y 轴、Z 轴坐标列表	基于CGCS2000 地心坐标系

变量	符号	变量描述	注释
输入变量	$ \left[ {R,\theta ,\varepsilon } \right] $	原点距离、方位角、俯仰角列表	基于极坐标系
输出变量	$ \left[ {L,B,H} \right] $	经度、纬度、高度列表	基于WGS-84坐标系

变量	符号	变量描述	注释
输入变量	${{\boldsymbol{\hat x}}^{{i}}}$	传感器$i$的状态估计	基于CGCS2000 地心坐标系
	${{\boldsymbol{P}}^{{i}}}$	传感器$i$的误差协方差	−
	${{\boldsymbol{\hat x}}^{{j}}}$	传感器$j$的状态估计	基于CGCS2000 地心坐标系
	${{\boldsymbol{P}}^{{j}}}$	传感器$j$的误差协方差	−
输出变量	${\boldsymbol{\hat x}}$	融合后的状态估计	基于CGCS2000 地心坐标系
输出变量	${\boldsymbol{P}}$	融合后的误差协方差	−

变量	符号	变量描述	注释
输入变量	${{\boldsymbol{P}}_{{{ij}}}}$	传感器$i$和传感器$j$的互协方差	${{\boldsymbol{P}}_{{{ij}}}} ={\boldsymbol{ }}{{\boldsymbol{P}}_{{{ji}}}}$
输出变量	$ {{\boldsymbol{D}}_{{{ij}}}} $	传感器$i$和传感器$j$的估计误差	${{\boldsymbol{D}}_{{{ij}}}} ={{\boldsymbol{D}}_{{{ji}}}}$

变量	符号	变量描述	注释
输入变量	$\hat {\boldsymbol{x}}_{k/k}^i$	传感器$i$在k时刻的状态估计	−
	$\hat {\boldsymbol{x}}_{k/k-1}^i$	传感器i在k时刻的状态预测	基于上一时刻
	${\boldsymbol{P}}_{k/k-1}^i$	传感器$i$在k时刻的估计协方差	−
	$\hat {\boldsymbol{x}}_{k/k-1}$	系统航迹在k时刻的状态预测	基于上一时刻
	$ {\boldsymbol{P}}_{k/k-1}$	系统航迹在k时刻的预测协方差	基于上一时刻
输出变量	$\hat {\boldsymbol{x}}_{k/k}$	融合后在k时刻的状态估计	−
输出变量	$ {\boldsymbol{P}}_{k/k}$	融合后在k时刻的误差协方差	−