一种冲击噪声下的多目标跟踪算法  

A multi-target tracking algorithm under the impulsive noise

作　　者：VU Van Toi 高洪元[1] 孙溶辰 陈暄 VU Van Toi;GAO Hongyuan;SUN Rongchen;CHEN Xuan(College of Information and Communication Engineering,Harbin Engineering University,Harbin 150001,China)

机构地区：[1]哈尔滨工程大学信息与通信工程学院,黑龙江哈尔滨150001

出　　处：《应用科技》2024年第1期130-135,142,共7页Applied Science and Technology

基　　金：国家自然科学基金项目(62073093,61571149);黑龙江省自然科学基金项目(LH2020F017)。

摘　　要：针对现有的子空间类多目标跟踪算法无法对相干目标进行有效跟踪,传统的动态跟踪方法在冲击噪声环境下失效的问题,提出了一种冲击噪声下的多目标跟踪算法。构造了一种新的零记忆非线性处理方法实现去冲击,推导得到了基于协方差矩阵更新的极大似然多目标跟踪方程,并设计了一种量子猫群算法,对其进行快速准确求解,实现了在恶劣噪声环境下的鲁棒多目标跟踪。仿真结果表明,所设计的算法突破了已有跟踪方法的性能和应用局限。本文分析结果可用于指导被动雷达和感知系统的跟踪模块设计。A new real-time multi-target tracking algorithm is proposed to address the problems that the existing subspace tracking algorithms cannot track coherent targets effectively and the traditional adaptive tracking methods fail in an impulsive noise environment.A new zero-memory nonlinear processing method is constructed to eliminate the impact,and the maximum likelihood multi-target tracking equation based on the covariance matrix updating is derived.A quantum cat swarm algorithm is designed to solve the multi-target tracking equation quickly and accurately,achieving robust multi-target tracking in a harsh noise environment.Simulation results show that the designed algorithm breaks through the performance and application limitation of existing tracking methods,and the analysis result can be applied to the guidance in the design of tracking module of passive radar and sensing system.

关 键 词：角度跟踪 冲击噪声 方位角估计 阵列测向 演化计算 量子群智能 智能优化算法 猫群优化算法 

分 类 号：TN911.7[电子电信—通信与信息系统]