【摘要】 在雷达跟踪系统中 ,广泛使用混合坐标系下扩展卡尔曼滤波器 (EKF)和转换卡尔曼滤波器 (CMKF)两种算法。但是当目标距离较远时 ,这两种滤波器由于量测方程非线性的影响 ,误差较大 ,甚至导致滤波发散。本文计算了传统 EKF滤波器引入的线性化误差 ,并在此基础上提出一种新的滤波算法 (NDRKF)。计算机仿真结果表明 ,该算法减小了线性化误差的影响 ,较明显地改善了滤波器的性能。更多还原

【Abstract】 In radar object tracking system,EKF and EMKF are wildly used for the nonlinearity of radar measurement aquation.However,when the target range is large enough the performance of the EKF results in considerable errors and even divergence due to the nonlinear effect.This paper first estimates the linearization error caused by EKF,and then develops a new nonlinear filter algorithm NDRKF,which is compared with EKF and CMKF D in computer simulations showing that the new filter considerably raduces the influence of the linearization error and track the target more precisely.更多还原