组合导航中一种新息自适应卡尔曼滤波算法

全球定位系统/航位推算组合导航定位中,由于目标运动的不确定性,GPS接收机与DR器件接收的数据存在噪声,使预置目标运动模型通常很难得到较高跟踪精度,针对应用常规卡尔曼滤波进行组合导航解算由于噪声统计特性未知而引起滤波不稳定的问题,本文提出了一种基于新息序列的量测计算进行自适应估计的卡尔曼滤波算法。该算法通过对新息方差强度进行极大似然估计,将新息计算引入卡尔曼滤波器的增益计算,达到控制发散的目的。最后对改进的算法与一般卡尔曼滤波算法做了对比仿真试验分析,结果表明了改进算法的有效性。

Application of New Information Adaptive Kalman Filter in Integrated Navigation

In the GPS/DR integrated navigation system, Noise existed in GPS receiver and DR device due to the uncertain moving objects. It is difficult to obtain high tracking accu- racy under a presetting target motion model. A Kalman filter algorithm based on Innovation Adaptive Estimation （IAE） is proposed in this paper aiming at figure out unstable filtering in conventional Kalman filtering caused by Unknown noise covariance. The algorithm introduces Innovation into the gain calculation in kalman filtering through maximum likelihood estimation of Innovation variance intensity in order to control divergence. In the end, practical test on the new algorithm and the conventional Kalman filter algorithm comparably is given, the results show that the effectiveness of the improved algorithm.