闪烁噪声下的自适应无迹卡尔曼算法

Adaptive Unscented Kalman Algorithm for Flicker Noise

针对非高斯噪声环境下过程噪声统计特性未知时机动目标跟踪精度不高甚至发散的问题,提出了一种应用EM算法和无迹卡尔曼滤波(Unscented Kalman filter,UKF)相结合的方法。该方法借助EM算法估计出较准确的过程噪声参数,再使用无迹卡尔曼滤波算法获得高精度的目标运动状态。仿真实验结果表明,该方法可以有效抑制滤波发散并显著提高跟踪精度。

Aiming at the problem that the moving target tracking accuracy is not high or even divergence when the process noise is not known under the condition of non-Gaussian noise, a method combining EM algorithm with unscented kalman filter is proposed. The method estimates more accurate process noise parameters by using the EM algorithm, and then uses the unscented kalman filter algorithm to obtain a high-precision target motion state. The simulation results show that this method can effectively suppress the filter divergence and improve the tracking ac-curacy significantly.