摘要
在以到达角(AoA)和到达时间(ToA)作为观测量的目标跟踪中,已有的非线性卡尔曼滤波很难同时兼顾低计算复杂度和高跟踪精度.针对该问题,提出了一种简单有效的偏差补偿卡尔曼滤波算法(BCKF).该算法首先利用AoA和ToA的等价几何关系对非线性观测方程进行伪线性化,得到伪线性卡尔曼滤波(PLKF).为解决PLKF面临的偏差问题,BCKF从理论推导了偏差构成项以及分析各项偏差的影响,然后补偿由伪线性观测噪声向量和观测矩阵相关性引起的偏差,实现更准确的状态估计.复杂度的理论分析和滤波性能的仿真结果表明,相较于其他滤波算法,BCKF以较低的计算开销实现了更高的跟踪精度,低噪声情况下可达到后验克拉美罗下界,并且收敛速度快,不容易受到初始误差的影响.
In target tracking with angle-of-arrival( AoA) and time-of-arrival( ToA) measurements,existing nonlinear Kalman filter algorithms cannot achieve both low computational complexity and high tracking accuracy. To address this problem,in this paper,we propose a simple and effective algorithm called the bias-compensation Kalman filter( BCKF). First,nonlinear measurement equations are pseudo-linearized by utilizing the equivalent geometric relationship between the AoA and ToA,which yields a pseudo-linear Kalman filter( PLKF).To overcome the bias associated with the PLKF,a detailed theoretical bias expression is derived,and the effect of each bias term is analyzed. Then,the BCKF compensates for the bias caused by the correlation of the pseudo-linear measurement of the noise vector and the measurement matrix to achieve more accurate state estimation. A theoretical analysis of the complexity and the simulation results of the filter performance show that BCKF achieves higher tracking accuracy with lower computational overhead compared with other filter algorithms,and attains the posterior Cramér-Rao lower bound over the mild noise region. Furthermore,BCKF enables fast convergence and is not susceptible to initial errors.
作者
林杰
齐望东
刘鹏
赵跃新
LIN Jie;QI Wangdong;LIU Peng;ZHAO Yuexin(Command,and Control Engineering College,Army Engineering University,Nanjing 210007,China;School of Information Science and Engineerings Southeast University,Nanjing 210096,China;Purple Mountain Laboratory,Nanjing 211111,China)
出处
《信息与控制》
CSCD
北大核心
2020年第6期657-666,共10页
Information and Control
基金
国家自然科学基金资助项目(61402520,61573376)。
关键词
目标跟踪
卡尔曼滤波
伪线性估计
偏差补偿
target tracking
Kalman filter
pseudo linear estimation
bias-compensation
