为利用无源固定单站对运动辐射源快速定位,将不敏卡尔曼滤波(UKF)和后向平滑算法应用于单站无源定位,给出了一种基于后向平滑的改进UKF算法;该算法利用当前时刻滤波结果通过后向平滑算法平滑估计前一时刻状态向量和协方差矩阵的估计值,为前向滤波提供较高精度的起始值。仿真结果表明改进的UKF算法在保证实时性的基础上改善了定位性能。
To achieve fast location of moving emitter by a non-moving single passive observer, an improved UKF algorithm based on backward-smoothing is proposed. The proposed algorithm smoothes the previous state vector and covariance matrix by the backward-smoothing algorithm using the current filtering results. Then, an initial value of higher precision is obtained. Simulation results indicate that the improved UKF algorithm may improve the location performance while keeping the real-time characteristic.