随着海上力量的快速兴起，新型快艇、舰船、无人机等设备不断涌现，海面目标类型更加丰富，精确导航定位是船只航行的必要辅助方式。然而，海面杂波及海上障碍物对目标跟踪带来不同程度的干扰，导致目标状态估计精度低及船只流量错误统计。因此，开展导航雷达抑制杂波干扰的跟踪算法研究具有十分重要的意义。

针对单部导航雷达在点目标跟踪算法的滤波及关联时所出现的断航，杂波引起虚假航迹的问题，提出基于量测转换的扩展目标跟踪算法。通过分析导航雷达采集到的目标信息特征，利用量测转换滤波，将目标点迹无偏转换到笛卡尔坐标系中，采用密度聚类方法实现目标点迹的凝聚处理，基于随机矩阵的扩展目标跟踪算法，对各个已跟踪的航迹进行预测，结合最近邻的目标关联算法，将落入相应波门内的目标分配给已有航迹，并更新目标航迹，实现扩展目标的大小、形状及航向上的跟踪。基于NAVICO导航雷达对海面航行船只的实测数据，相较于点目标跟踪算法，扩展多目标跟踪航迹准确度提高45.8%，并提供更多的目标运动特征信息。

搭建雷达组网测试平台，通过误差匹配、坐标变换、时间配准的流程对两部导航雷达测量的点迹预处理，将目标点迹与目标航迹关联，多雷达点迹数据合并处理，基于贝叶斯最优解进行量测分配及更新航迹，得到目标的运动状态。相较于单部雷达，组网雷达多目标跟踪的最优子模式分配距离均值下降48.48%，实现对目标高精度的扩展信息估计和更稳定的跟踪。

实测结果表明，改进的算法及组网雷达信息融合能够较好的改善目标跟踪精度，为海面船只在辅助导航领域提供一定的参考价值。

With the rapid rise of maritime forces, new types of speedboats, ships, unmanned aerial vehicles and other equipment are emerging, and the types of surface targets are more abundant. Accurate navigation and positioning is a necessary auxiliary way for ships to navigate. However, sea clutter and obstacles interfere with target tracking to varying degrees, resulting in low accuracy of target state estimation and incorrect ship flow statistics. Therefore, it is very important to study the tracking algorithm of navigation radar to suppress clutter interference.

An extended target tracking algorithm based on measurement conversion is presented to solve the problem of false tracks caused by clutter when point target tracking algorithm is filtered and correlated by a single navigation radar. By analyzing the characteristics of target information collected by navigation radar, the target points are transformed unbiased into Cartesian coordinate system using measurement conversion filter, the target points are aggregated by density clustering method, and each tracked track is predicted by extended target tracking algorithm based on random matrix. Targets falling into the corresponding gate are assigned to the existing track by combining the nearest neighbor target association algorithm. The target track is updated to track the size, shape and heading of the extended target. Compared with point target tracking algorithm, extended multi-target tracking track accuracy improves 45.8% based on NAVICO navigation radar's measured data for sea-going vessels. And provide more target motion feature information.

Radar networking test platform is set up. Through the process of error matching, coordinate transformation and time registration, the points and tracks of two navigation radar measurements are preprocessed, the target points and tracks are associated, the multi-radar points and tracks are merged and processed, and the track is measured, distributed and updated based on the Bayesian optimal solution to get the moving state of the target. Compared with single radar, the average value of optimal sub-pattern assignment range mean of netted radar for multi-target tracking is reduced by 48.48%, which achieves high-precision extended information estimation and more stable tracking.

The actual measurement results show that the improved algorithm and network radar information fusion can effectively improve target tracking accuracy, providing certain reference value for sea vessels in the field of auxiliary navigation.

[1] Katzmann Alexander, Taubmann Oliver, Ahmad Stephen, et al. Explaining clinical decision support systems in medical imaging using cycle-consistent activation maximization[J]. Neurocomputing, 2021, 458: 141-156.

[2] Hernangómez Rodrigo, Visentin Tristan, Servadei Lorenzo, et al. Improving Radar Human Activity Classification Using Synthetic Data with Image Transformation[J]. Sensors, 2022, 22(4): 1519-1519.

[3] Koch J.W.. Bayesian approach to extended object and cluster tracking using random matrices[J]. IEEE Transactions on Aerospace and Electronic Systems, 2008, 44(3): 1042-1059.

[4] Jian Lan, X. Rong Li. Extended-Object or Group-Target Tracking Using Random Matrix With Nonlinear Measurements[J]. IEEE Trans. Signal Processing, 2019, 67(19): 5130-5142.

[5] Erich Schubert, Jörg Sander, Martin Ester, Hans Peter Kriegel, et al. DBSCAN Revisited, Revisited[J]. ACM Transactions on Database Systems (TODS), 2017, 42(3): 1-21.

[6] Dong Wenhao, Wang Jingyi, Song Zhiyong. Mono-pulse radar angle estimation algorithm under low signal-to-noise ratio[J]. Journal of Physics: Conference Series, 2021, 1914(1).

[7] Y. Guo, Y. Li, R. Tharmarasa, et al. GP-PDA Filter for Extended Target Tracking With Measurement Origin Uncertainty[J]. IEEE Transactions on Aerospace and Electronic Systems, 2019, 55(4): 1725-1742.

[8] Guo Yunfei, Yang Dongsheng, Ren Lei, et al. Gaussian processes based extended target tracking in polar coordinates with input uncertainty[J]. EURASIP Journal on Advances in Signal Processing, 2022, 2022(1).

[9] Karl Granström, Umut Orguner. New prediction for extended targets with random matrices[J]. IEEE Trans. Aerospace and Electronic Systems, 2014, 50(2): 1577-1589.

[10] Sun Lifan, Zhang Jinjin, Yu Haofang, et al. Maneuvering Extended Object Tracking with Modified Star-Convex Random Hypersurface Model Based on Minimum Cosine Distance[J]. Remote Sensing, 2022, 14(17): 4376.

[11] Liu Shun, Liang Yan, Xu Linfeng. Maneuvering extended object tracking based on constrained expectation maximization[J]. Signal Processing, 2022, 201.

[12] Liu Jiaqi, Wang Zhen, Cheng Di, et al. Marine Extended Target Tracking for Scanning Radar Data Using Correlation Filter and Bayes Filter Jointly[J]. Remote Sensing, 2022, 14(23): 5937-5937.

[13] Zhang Xing, Yan Zhibin, Chen Yunqi, et al. A novel particle filter for extended target tracking with random hypersurface model[J]. Applied Mathematics and Computation, 2022, 425.

[14] Steven V. Bordonaro, Peter Willett 0001, Yaakov Bar-Shalom. Decorrelated unbiased converted measurement Kalman filter[J]. IEEE Aerospace and Electronic Systems, 2014, 50(2): 1431-1444.

[15] 姜佰辰, 关键, 周伟, 等. 基于多项式卡尔曼滤波的船舶轨迹预测算法[J]. 信号处理, 2019, 35(05): 741-746.

[16] 卜瑞波, 徐海祥, 冯辉, 等. 一种面向船舶智能航行的海上目标实时跟踪算法[J]. 大连理工大学学报,2020,60(01):30-35.

[17] 矫腾章, 胡玉新, 吕鹏, 等. 一种在轨海上多运动舰船目标检测和跟踪方法[J]. 中国科学院大学学报,2020, 37(03): 368-378.

[18] 许红, 谢文冲, 高飞, 等. 机载预警雷达海面多目标航迹起始算法研究[J]. 雷达科学与技术, 2015, 13(06): 655-659+666.

[19] 陈辉, 魏凤旗, 韩崇昭. 多扩展目标跟踪优化中基于威胁规避的无人机路径规划策略[J/OL]. 雷达学报: 1-12.

[20] 任磊, 郭云飞. 带输入噪声的高斯过程扩展目标跟踪算法[J]. 杭州电子科技大学学报(自然科学版), 2022, 42(02): 41-48+95.

[21] 梁苑, 戚国庆, 陈烨, 等. 不完全量测下事件触发水面扩展目标跟踪[J/OL]. 兵工学报: 1-11.

[22] 陈振, 李翠芸, 李想. B样条曲面三维扩展目标跟踪算法[J/OL]. 西安电子科技大学学报: 1-12.

[23] 陈辉, 魏凤旗, 韩崇昭. 多扩展目标跟踪中基于多特征优化的传感器控制方法[J]. 电子与信息学报, 2023,45(01): 191-199.

[24] 于泽祥, 蔡宗平, 杨剑, 等. IMM-SCKF算法在海面扩展目标跟踪中的应用[J]. 现代防御技术, 2019, 47(04): 90-96.

[25] 李冬冬, 范红旗, 王丽萍, 等. 基于相关滤波器的雷达单扩展目标跟踪[J]. 航空兵器, 2022, 29(02): 19-23.

[26] 李国财. 随机超曲面模型容积卡尔曼扩展目标跟踪算法[J]. 舰船电子工程, 2021, 41(02): 31-35.

[27] Lei Chai, Lingjiang Kong, Suqi Li, et al. The multiple model multi-Bernoulli filter based track-before-detect using a likelihood based adaptive birth distribution[J]. Signal Processing, 2020, 171(C): 107501-107501.

[28] Chiman Kwan, Bence Budavari. A high-performance approach to detecting small targets in long-range low-quality infrared videos[J]. Signal, Image and Video Processing, 2021, 16: 93-101.

[29] Shuo Liu, Hongbo Li, Yun Zhang, et al. Multiple hypothesis method for tracking move-stop-move target[J]. The Journal of Engineering, 2019, 2019(19): 6155-6159.

[30] Lerro D., Bar-Shalom Y.. Tracking with debiased consistent converted measurements versus EKF[J]. IEEE Transactions on Aerospace and Electronic Systems, 1993, 29(3): 1015-1022.

[31] Sen Wang, Qinglong Bao. Decorrelated unbiased converted measurement for bistatic radar tracking[J]. Journal of Applied Remote Sensing, 2021, 15(1).

[32] Guoyong Wang, Xiaoliang Feng. Unbiased converted measurement manoeuvering target tracking under maximum correntropy criterion[J]. Cognitive Computation and Systems, 2020, 2(3): 125–129.

[33] 刘福, 马保东, 马祥, 等. 一种基于多模型量测转换的机动目标跟踪算法[J].舰船电子工程, 2022, 42(05): 36-39+95.

[34] 郭云飞, 任昕. 一种极坐标下基于高斯过程的扩展目标跟踪方法[J]. 控制理论与应用, 2020, 37(07): 1501-1510.

[35] 吴麒, 熊杰, 宁静. 一种量测转换的移动外辐射源多站协同跟踪算法[J]. 电讯技术, 2020, 60(09): 1087-1095.

[36] 王春梅. 基于极坐标的CMKF定位算法研究[J]. 信息技术, 2014(06): 28-29+33.

[37] Lu Xiujuan, Yi Wei, Wang Yao. Adaptive beam burden management for multi-target tracking in distributed radar network[J]. Signal Processing, 2022, 200.

[38] Shui Yongtao, Wang Yu, Li Yu, et al. Consensus-Based Distributed Target Tracking with False Data Injection Attacks over Radar Network[J]. Applied Sciences, 2021, 11(10): 4564-4564.

[39] He Bin, Su Hongtao, Huang Junsheng. Joint Beamforming and Power Allocation Between A Multistatic MIMO Radar Network and Multiple Targets Using Game Theoretic Analysis[J]. Digital Signal Processing, 2021(prepublish): 103085.

[40] 路俊维, 马雪松, 褚建立. 高速船舶雷达动态组网中避障恢复算法[J]. 舰船科学技术, 2021, 43(14): 139-141.

[41] 何子述, 苏洋, 程婷. 基于共址MIMO雷达组网系统的机会协同目标跟踪方法[J]. 空天防御, 2022, 5(01): 6-11.

[42] 韩晓斐, 何华锋, 何耀民, 等. 主被动异构弹载雷达组网抑制假目标干扰方法[J]. 兵工学报, 2022, 43(12): 3142-3150.

[43] 陈敬军. 一种基于目标点迹关联的舰载多雷达数据融合系统[J]. 雷达与对抗, 2021, 41(03): 28-32+42.

[44] 高畅, 谷丰登, 严俊坤, 等. 多普勒信息辅助的网络化雷达融合检测[J/OL]. 雷达学报: 1-16.

[45] 时晨光, 董璟, 周建江. 频谱共存下面向多目标跟踪的组网雷达功率时间联合优化算法[J/OL]. 雷达学报: 1-12.

[46] 曾雅俊, 王俊, 魏少明, 等. 分布式多传感器多目标跟踪方法综述[J]. 雷达学报, 2023, 12(01): 197-213.

[47] 甄绪, 刘方. 基于改进的FCM和信息熵修正的航迹融合算法[J]. 航空学报, 2022, 43(05): 420-431.

[48] 施治国, 熊文芳. 时变定位误差的分布式雷达组网航迹融合算法[J]. 电子测量技术, 2018, 41(11): 20-25.

[49] 李月武, 胡建旺, 吉兵. 一种新的航迹融合算法精度评价指标[J]. 西安电子科技大学学报, 2020, 47(01): 111-119.

[50] 陈永利. 多船舶雷达信息的航迹算法与目标跟踪研究[J]. 舰船科学技术, 2021, 43(16): 40-42.

[51] 韩晓斐, 何华锋, 何耀民. 主被动异构弹载雷达组网抑制假目标干扰方法[J]. 兵工学报, 2022, 43(12): 3142-3150.

[52] 蒋志强, 刘思成. 干扰条件下的雷达组网包络三维融合方法[J]. 现代计算机, 2022, 28(13): 32-38.

[53] 郭立新, 魏仪文. 复杂动态海面与目标电磁散射及回波仿真研究现状与展望[J]. 雷达学报, 2023, 12(01): 76-109.

[54] 张玉石, 李笑宇, 张金鹏, 等. 基于深度学习的海杂波谱参数预测与影响因素分析[J]. 雷达学报, 2023, 12(01): 110-119.