基于IMM-SCKF的海上机动目标跟踪算法研究  

Maritime Maneuvering Target Tracking Using IMM-SCKF Algorithm

作　　者：李思嘉 曹菲 林浩申 LI Si - jia1,2, CAO Fei1, LIN Hao - shen1(1. Rocket Force University of Engineering, Xi＇an Shanxi 710025, China; 2. Rocket Force Military Representative Office in Factory 7103, Xi＇an Shanxi 710100, China)

机构地区：[1]火箭军工程大学,陕西西安710025 [2]火箭军驻7103厂军事代表室,陕西西安710100

出　　处：《计算机仿真》2018年第10期288-294,共7页Computer Simulation

摘　　要：以海上机动目标无源跟踪为背景,构建了基于弹载探测平台的无源跟踪系统。针对目标强机动性和跟踪系统强非线性的特点,在交互式多模型滤波(IMM)的基础上,引入基于贝叶斯估计的SCKF算法。通过推导IMM算法中协方差矩阵在SCKF算法中的等价描述,解决了两种算法输入输出不匹配的问题,并设计了IMM-SCKF算法的算法流程。仿真结果表明,上述算法能快速适应目标的机动行为,准确估计目标运动模型的概率,实现对机动目标的高精度跟踪,其位置跟踪平均精度和速度跟踪平均精度分别达到53. 79m和1. 54m/s,验证了该算法的有效性。Based on monitoring and tracking the maritime maneuvering target, the passive tracking system of missile observation platform is constructed. Due to the strong mobility of target and strong nonlinearity of tracking system, an algorithm based on Bayes estimation is introduced in the interacting multiple model （IMM）. By deriving the equivalent description of IMM＇s covariance matrix in SCKF, this paper solves the problem of mismatch between the input and output of the two algorithms, and the IMM - SCKF algorithm flowchart is designed. The simulation results show that the proposed algorithm can adapt to the maneuverability quickly and estimate the probability of target motion model accurately. The average position and speed tracking accuracy can reach 53.79mand 1.54m/srespectively, which verifies the effectiveness of the proposed algorithm.

关 键 词：机动目标跟踪 弹载平台 交互式多模型 平方根容积卡尔曼滤波 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]