﻿ 跟踪反馈技术在舰载雷达目标检测的应用
 舰船科学技术  2023, Vol. 45 Issue (11): 127-130    DOI: 10.3404/j.issn.1672-7619.2023.11.025 PDF

Application of tracking feedback technology in shipborne radar target detection method
LI Hui-shu
Shanxi Institute of Energy, Jinzhong 030600, China
Abstract: The emergence of phased array radar enables shipboard radar to change the beam direction by controlling the transmit and receive phase difference between the array elements, which improves the scanning speed of the radar beam. However, the multi-target tracking algorithm based on RFS modeling still faces the problem of multi-target tracking accuracy in nonlinear systems. How to effectively learn from and use nonlinear approximation methods in single-target filtering algorithms to improve shipborne radar tracking feedback technology. Accordingly, this paper will focus on the application of tracking feedback technology in shipborne radar target detection methods, and carry out relevant analysis and research with δ-GLMB tracking feedback algorithm. It aims to use the more comprehensive and deeper dimensional information provided by multiple sensors to improve the observation accuracy of the entire tracking feedback technology system, thereby enhancing the anti-false detection, missed detection, and false detection capabilities of the shipboard radar.
Key words: tracking feedback technology     shipborne radar     object detection
0 引　言

1 δ-GLMB算法介绍 1.1 δ-GLMB的一般实现方法

 $\pi(X)=\Delta(X) \sum_{(I, g) \in \mathcal{F}(\mathrm{I}) \times \mathrm{Z}} w^{(I, s)} \delta_{I}(\mathcal{L}(X))\left[p^{g}\right]^{X}，$
 $\Delta(X)=\left\{\begin{array}{ll} 1, & \operatorname{card}(X)=\operatorname{card}(\mathcal{L}(X))，\\ 0, & \operatorname{card}(X) \neq \operatorname{card}(\mathcal{L}(X))。\end{array}\right.$

 $\pi_+\left(X_+\right)=\Delta\left(X_+\right) \sum_{(L, g)=F\left(\mathrm{~L}_+\right)=\mathrm{B}} w_+^{(L, g)} \delta_{L_+}\left(\mathcal{L}\left(X_+\right)\right)\left[p_+^{(g)}\right]^{X} 。$

 $w_{\gamma}(L)=\prod_{\ell \in \mathbb{B}}\left(1-\varepsilon_{\gamma}^{(\rho)}\right) \prod_{l \in \mathbb{B}} \frac{1_{\mathbb{B}}(\ell) \varepsilon_{\gamma}^{(l)}}{1-\varepsilon_{r}^{(\ell)}} 。$

1.2 δ-GLMB的序贯蒙特卡罗实现

1）预测步骤

 $p^{(\vartheta)}(\cdot, \ell)=\left\{w_{i}^{(\vartheta)}(\ell), {\boldsymbol{x}}_{i}^{(\vartheta)}(\ell)\right\}_{i=1}^{J_{\gamma=1}^{(\vartheta)}(\ell))} ，$
 $p_{\gamma}^{(\vartheta)}(\cdot, \ell)=\left\{w_{\gamma, i}^{(\vartheta)}(\ell), {\boldsymbol{x}}_{\gamma, i}^{(\vartheta)}(\ell)\right\}_{i=1}^{J_{\gamma}^{(\vartheta)}}(\ell))。$

 $p^{(\vartheta, \theta)}(\cdot, \ell \mid Z)=\left\{\frac{\varphi_{z}\left(x_{n}^{(\vartheta)}(\ell), \ell ; \theta\right) w_{n}^{(\vartheta)}(\ell)}{\eta_{z}^{(\vartheta, \theta)}(\ell)}, {\boldsymbol{x}}_{n}^{(\vartheta)}(\ell)\right\}_{n-1}^{J^{(\vartheta)}(\ell)} 。$

2 基于SICKF的δ-GLMB跟踪目标识别方法

 图 1 SICKF-δ-GLMB算法流程框图 Fig. 1 Flow Chart of SICKF-GLMB Algorithm
3 分布式舰载雷达信息融合的MM-SICKF-δ-GLMB算法及目标检测

3.1 MM-SICKF-δ-GLMB算法实现

 $\pi(X)=\Delta(X) \sum_{c \in \mathbb{C}} w^{(c)}(\mathcal{L}(X))\left[p^{(c)}\right]^{X} 。$

 $\pi_+\left(\mathrm{X}_+\right)=\Delta\left(\mathrm{X}_+\right) \sum_{c \in \mathbb{C}} w_+^{(c)} \mathcal{L}\left(\mathrm{X}_+\right)\left[p_+^{(c)}\right]^{\mathrm{x}_+} 。$

 $\pi(X \mid Z)=\Delta(X) \sum_{c \in \mathbb{C} \theta \in \Theta} w_{z}^{(c, \theta)}(\mathcal{L}(X))\left[p^{(c, \theta)}(-\mid Z)\right]^{X}。$
3.2 基于时空双重CI方法的MM-SICKF-δ-GLMB算法实现

k时刻舰载雷达传感器的量测集合如下：

 ${{{\boldsymbol{Z}}}}^i_{Radar,k}=[Z^1_{Radar,k},Z^2_{Radar,k},\cdots,Z^M_{Radar,k}]。$

 图 2 不同速度下的目标识别漂移量随时间变化曲线 Fig. 2 Variation curve of target recognition drift with time at different speeds
 $Z_{k}^{i}=\left[Z_{k}^{1}, Z_{k}^{2}, \cdots, Z_{k}^{N_{1}}\right] 。$
3.3 仿真实验结果分析

 图 3 运动加速度随时间变化曲线 Fig. 3 Motion acceleration versus time curve

4 结　语

 [1] 李兵舰, 冯丁舜, 张殿友. 舰载雷达干扰设备多目标干扰能力分析[J]. 舰船电子对抗, 2019, 42(1): 23-28+33. [2] 王宇, 魏永峰. 舰载雷达系统的分布式模拟目标相关性分析[J]. 电子信息对抗技术, 2010, 25(3): 61-64. [3] 钟娟, 刘以安, 吴少鹏, 等. 舰载雷达低空目标回波信号仿真研究[J]. 雷达与对抗, 2004(3): 4-5. [4] 施端阳, 林强, 胡冰, 等. 深度学习在雷达目标检测中的应用综述[J]. 雷达科学与技术, 2022, 20(6): 17-18. [5] 安金坤, 方峰, 曾巍. 一种基于双测度的参数重叠舰载雷达型号自动识别方法[J]. 电子信息对抗技术, 2022, 37(3): 16-18. [6] 卢锦, 马令坤, 吕春玲, 等. 基于代价参考粒子滤波器组的多目标检测前跟踪算法[J]. 自动化学报, 2023, 49: 10-11. [7] 刘衍会, 张宏武. 强杂波下微弱目标检测方法研究[J]. 现代雷达, 2022, 44(7): 31-37.