针对现有粒子滤波在港口船舶跟踪中受地理边界、速度、加速度和安全距离等动态约束影响导致跟踪精度下降问题，提出一种含动态多约束的辅助粒子滤波（Dynamic Multi-Constraint Auxiliary Particle Filter，DMCAPF）。首先，将上述物理约束转化为伪测量值，通过违反度量与惩罚函数搭建软约束机制，并嵌入贝叶斯框架。其次，设计双层自适应约束权重框架：第一层采用三类敏感函数生成基础权重，第二层利用随机森林模型融入环境特征实现动态修正。最后，将优化的约束权重嵌入辅助粒子滤波采样两阶段，修正粒子选择与权重更新，提高约束粒子的传播概率。实验结果表明，DMCAPF的跟踪精度可达9.8 m，约束满足率为95.4%，计算时间为0.25 s，适用于港口船舶跟踪。
Existing particle filters for port ship tracking suffer from accuracy degradation due to dynamic constraints such as geographical boundaries, velocity, acceleration, and safety distance. To address the issue, this paper proposes a Dynamic Multi-Constraint Auxiliary Particle Filter (DMCAPF), which transforms physical constraints into pseudo-measurements and establishes a soft constraint mechanism via violation metrics and penalty functions embedded in the Bayesian framework. To deal with multi-constraint conflicts and sensitivity, a two-layer adaptive constraint weight framework is designed, where the first layer generates basic weights using three sensitivity functions and the second layer performs dynamic correction through a random forest model incorporating environmental features. The optimized constraint weights are then integrated into two-stage sampling in auxiliary particle filter to modify particle selection and weight update, which improves the propagation probability of constraint-satisfying particles. Experimental results show that tracking accuracy is 9.8 m, constraint satisfaction rate is 95.4%, and computation time is 0.25 s of the proposed DMCAPF, demonstrating its suitability for port ship tracking.
2026,48(8): 167-174 收稿日期：2025-8-1
DOI：10.3404/j.issn.1672-7649.2026.08.026
分类号：U675.96；U664.82
基金项目：辽宁省教育厅基本科研项目（面上项目）（JYTMS20230862）
作者简介：李波(1977-),男,博士,教授,研究方向为现代信息处理技术
参考文献：
[1] XU L, LI X R, LIANG Y, et al. Constrained dynamic systems: generalized modeling and state estimation[J]. IEEE Transactions on Aerospace and Electronic Systems, 2017, 53(5): 2594-2609
[2] ARULAMPALAM M S, MASKELL S, GORDON N, et al. A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking[J]. IEEE Transactions on Signal Processing, 2002, 50(2): 174-188
[3] 薛锋, 刘忠, 石章松. 基于粒子滤波的约束目标被动跟踪研究[J]. 武汉理工大学学报(交通科学与工程版), 2007(1): 43-45+52 XUE F, LIU Z, SHI Z S. Research on the passive tracking of constrained targets using particle filtering[J]. Journal of Wuhan University of Technology (Transportation Science & Engineering), 2007(1): 43-45+52
[4] GONG Z, GAO G, WANG M. An adaptive particle filter for target tracking based on double space-resampling[J]. IEEE Access, 2021, 9: 91053-91061
[5] ZHANG H W, XIE W X. Constrained auxiliary particle filtering for bearings-only maneuvering target tracking[J]. Journal of Systems Engineering and Electronics, 2019, 30(4): 684-695
[6] ZHANG H W, LI L G, XIE W X. Constrained multiple model particle filtering for bearings-only maneuvering target tracking[J]. IEEE Access, 2018, 6: 51721-51734
[7] YU H, FANG Z, MURRAY A T, et al. A direction-constrained space-time prism-based approach for quantifying possible multi-ship collision risks[J]. IEEE Transactions on Intelligent Transportation Systems, 2019, 22(1): 131-141
[8] MÜCKE N T, BOHTÉ S M, OOSTERLEE C W. The deep latent space particle filter for real-time data assimilation with uncertainty quantification[J]. Scientific Reports, 2024, 14(1): 19447
[9] TANG H, YIN Y, SHEN H. A model for vessel trajectory prediction based on long short-term memory neural network[J]. Journal of Marine Engineering & Technology, 2022, 21(3): 136-145
[10] MA L, GUO Z, SHI G. AIS data driven ship behavior modeling in fairways: a random forest based approach[J]. Applied Sciences, 2024, 14(18): 8484
[11] KARLSSON R, GUSTAFSSON F. Recursive Bayesian estimation: bearings-only applications[J]. IEEE Proceedings-Radar, Sonar and Navigation, 2005, 152(5): 305-313
[12] 叶翔. 受限水域船舶自适应神经网络轨迹跟踪控制方法研究[D]. 舟山: 浙江海洋大学, 2024.
[13] 夏鑫, 李宁, 刘鹏, 等. 基于随机森林和灰色关联法的目标威胁估[J]. 舰船科学技术, 2024, 46(13): 162-166 XIA X, LI N, LIU P, et al. Target threat assessment based on random forest and grey relational analysis[J]. Ship Science and Technology, 2024, 46(13): 162-166
[14] ZHANG W, XU G, SONG Y, et al. An obstacle avoidance strategy for complex obstacles based on artificial potential field method[J]. Journal of Field Robotics, 2023, 40(5): 1231-1244
[15] DU S, DENG Q. Unscented particle filter algorithm based on divide-and-conquer sampling for target tracking[J]. Sensors, 2021, 21(6): 2236
[16] XU C, WANG X, DUAN S, et al. Spatial-temporal constrained particle filter for cooperative target tracking[J]. Journal of Network and Computer Applications, 2021, 176: 102913
[17] CHENG Y, REN W, XIU C, et al. Improved particle filter algorithm for multi-target detection and tracking[J]. Sensors, 2024, 24(14): 4708
