自适应抗差UKF算法在GPS/BD-2组合系统中的应用

Application of adaptive robust UKF algorithm in GPS/BD-2 integrated system

针对非线性系统中因噪声模型不准确或测量数据中存在野值而导致无迹卡尔曼滤波(UKF)结果产生偏差甚至发散的问题,提出了一种自适应抗差无迹卡尔曼滤波(ARUKF)算法。该算法利用抗差估计原理构造抗差方差分量统计量,并由该统计量引入自适应因子来调节增益矩阵,减弱野值对滤波的影响。将ARUKF算法应用于GPS/BD-2组合导航系统中,仿真实验结果表明,当观测数据中存在野值时,ARUKF算法能够有效地控制观测异常误差的影响,定位精度得到了很大提高,并在不同系统噪声和观测噪声方差下,具有较好的鲁棒性和实时性。

In nonlinear system the unscented Kalman filter （UKF） resuhs often had error or even divergence when the error model was inaccurate or the measurement data contain out｜iers. So this paper introduced an adaptive robust unscented Kalman filter （ARUKF） algorithm to solve this problem by using robust estimation principle to introduce robust variance component statistics. It degraded the effects of outliers to filtering through adaptive factor based on the statistics to adjust gain matrix. Ap- plying the ARUKF algorithm to GPS/BD-2 integrated navigation system through simulating results show that ARUKF algorithm can effectively control influence of abnormal observations and improve the positioning accuracy when the measurement data contain outliers, and has good robustness and real-time under the different noise system and observation noise variances.