一种基于期望最大化去偏转换量测滤波的目标跟踪算法  

A target tracking algorithm based on expectation maximization debiased conversion measurement filter

作　　者：张连仲[1] 王宝宝 王超尘[2] ZHANG Lianzhong;WANG Baobao;WANG Chaochen(Jiangsu Automation Research Institute,Lianyungang 222006,China;School of Automation,Nanjing University of Science and Technology,Nanjing 210094,China)

机构地区：[1]江苏自动化研究所,连云港222006 [2]南京理工大学自动化学院,南京210094

出　　处：《中国惯性技术学报》2020年第6期829-833,共5页Journal of Chinese Inertial Technology

基　　金：航空科学基金(2016ZC59006)。

摘　　要：针对雷达测量系统因野值干扰导致跟踪精度下降的问题,提出了一种期望最大化去偏转换量测卡尔曼滤波(EMDCMKF)。先对目标量测去偏转换,将转换后的量测噪声协方差解耦合并适配自适应因子,之后利用期望最大化方法估计自适应因子矩阵,从而对量测噪声协方差进行修正。仿真结果表明,提出的EMDCMKF位置均方根误差相比于EKF和DCMKF分别减少了45.1%和52.5%,速度均方根误差分别减少了52.6%和66.8%,在野值干扰的环境下,所提出的EMDCMKF算法可以得到更高的估计精度。Aiming at the problem that the tracking accuracy of the radar measurement system decreases due to wild value interference,an Expectation Maximization Debiased Conversion Measurement Kalman Filter(EMDCMKF)is proposed.Firstly,the target measurement is debiased,the converted measurement noise covariance is decoupled and the adaptive factor is adapted.Then the expected maximization method is used to estimate the adaptive factor matrix,thus the measurement noise covariance is corrected.The simulation results show that the position root mean square error of the proposed EMDMCKF algorithm is reduced by 45.1%and 52.5%respectively,and the speed root mean square error is reduced by 52.6%and 66.8%respectively compared with EKF and DCCKF.The proposed EMDCMKF algorithm can obtain better estimation accuracy under the environment with wild value interference.

关 键 词：雷达测量系统 野值 期望最大化 去偏转换量测卡尔曼滤波 自适应因子 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]