自适应卡尔曼滤波在目标跟踪系统中的应用

目标跟踪是精确制导系统中的重要组成部分.文中针对运动目标跟踪问题,在建立运动模型的基础上,应用卡尔曼滤波算法进行了跟踪仿真研究.考虑到直角坐标系下的扩展卡尔曼滤波容易发散,可能导致滤波精度变差,所以文章提出一种针对非线性观测模型和线性动态模型的自适应推广卡尔曼滤波器.直角坐标系下的自适应卡尔曼滤波算法,对虚拟噪声进行了估计,动态补偿观测器模型的线性化误差,削减了系统的观测误差,并对其滤波理论及算法进行了仿真研究.结果表明:该算法提高了滤波的稳定性、快速性和精确性,优于一般的扩展卡尔曼滤波算法.

Application of Adaptive Extended Kalman Filter for Tracking a Moving Target

Object tracking is a very important part of precise guidance system.Aiming at moving target track problem,based on to building the moving model,the paper introduces Kalman filtering algorithm to do the research of tracking simulation.Taking into account the instability and low accuracy of passive filters in bearings-only target tracking,the paper presents an adaptive extended Kalman filter suited for nonlinear observation model and linear dynamic model.Virtual noise is estimated,and errors due to linearization are dynamically compensated so that the system's observation error is reduced.The filtering theory and the algorithm are studied.Simulation results show that MPAEKF can improve the filter convergence and accuracy.