| 基于叠加式传感器的多普勒雷达多目标联合检测与估计 |
| |
| 引用本文： | 董文豪,达凯,宋志勇,付强.基于叠加式传感器的多普勒雷达多目标联合检测与估计[J].信号处理,2022,38(5):964-972. |
| |
| 作者姓名： | 董文豪 达凯 宋志勇 付强 |
| |
| 作者单位： | 国防科技大学电子科学学院ATR国防重点实验室，湖南 长沙 410073 |
| |
| 基金项目： | 国家自然科学基金61401475 |
| |
| 摘 要： | 多目标检测与估计是多普勒雷达的基本任务。当信噪比较低时，为确保检测到目标需降低门限而产生了大量虚警，基于数据的多假设跟踪（Multi-Hypothesis Tracking， MHT）和联合概率数据关联（Joint Probabilistic Data Association， JPDA）方法因计算复杂度过高而失效，基于原始信号的随机有限集（Random Finite Set，RFS）滤波器可有效解决该问题。多普勒雷达回波信号以叠加的方式受到多个目标影响，其多目标检测与估计问题属于叠加式传感器的典型应用。本文在叠加式多伯努利（Multi-Bernoulli， MBR）滤波器基础上利用具有准确势估计的独立同分布群（Independent and Identically Distributed Cluster， IIDC）RFS对新生目标建模，并采用辅助粒子滤波器（Auxiliary Particle Filter， APF）实现了多目标联合检测与状态估计。仿真结果表明，混合MBR和集势概率假设密度（Cardinalized Probability Hypothesis Density，CPHD）滤波器对多普勒雷达多目标的检测估计性能优于MBR滤波器，且减小了计算复杂度。
|
| 关 键 词： | 叠加式传感器 联合检测与估计 多伯努利 集势概率假设密度 辅助粒子滤波 |
| 收稿时间： | 2021-09-01 |
Multi-target Joint Detection and Estimation of Doppler Radar Based on Superpositional Sensors |
| |
| Affiliation: | National Key Laboratory of Science and Technology on Automatic Target Recognition，College of Electronic Science，National University of Defense Technology，Changsha，Hunan 410073，China |
| |
| Abstract: | ? ?Multi-target detection and estimation is a basic task of Doppler radar. When the signal-to-noise ratio is low， in order to ensure that the target is detected， the threshold needs to be lowered and a large number of false alarms are generated. The multi-hypothesis tracking （MHT） and joint probabilistic data association （JPDA） methods based on data processing fail due to high computational complexity. Random finite set （RFS） filters based on original signal processing can effectively solve this problem. The Doppler radar echo signal is affected by multiple targets in a superpositional manner， and the problem of multi-target detection and estimation is a typical application of superpositional sensors. Based on the superpositional multi-Bernoulli （MBR） filter， this paper uses independent and identically distributed cluster （IIDC） RFS with accurate potential estimation to model the new target， and uses auxiliary particle filter The auxiliary particle filter （APF） realizes multi-target joint detection and state estimation. The simulation results show that the hybrid MBR and cardinalized probability hypothesis density （CPHD） filter has better performance for multi-target detection and estimation of Doppler radar than the MBR filter， and reduces the computational complexity. |
| |
| Keywords: | |
|
| 点击此处可从《信号处理》浏览原始摘要信息 |
|
点击此处可从《信号处理》下载全文 |
