低速旋转GNSS卫星导航双天线加权kalman定位测速方法

Study on the Weighted Kalman Positioning and Speed Measurement of Dual Antenna of Low-speed Rotating GNSS Satellite Navigation

GNSS常用于载体的定位测速。设想在旋转载体两侧对向各安装一个GNSS接收机天线,使得在任一历元下均能稳定接收GNSS信号。但当载体滚转时,GNSS观测量测量误差将使得传统kalman滤波器定位测速结果出现跳变。本文以低速旋转的用户模型为例,提出一种加权Kalman滤波算法。应用该方法的过程中,利用观测量残余与测量均方差的比值,调整超限观测量的方差,减小其滤波增益,达到弱化超限观测值的权重的目的,实现所有可见卫星均参与滤波,使得定位结果不会发生跳变。通过半实物仿真对本文所提出算法进行性能验证。从仿真实验结果可知,与传统的Kalman算法相比,本文算法的定位结果可用率(三维定位误差小于30 m)从89%提升至99%,三维测速误差(1σ)从0.19 m/s提升至0.11 m/s。

GNSS is widely used for the positioning and speed measurement of carriers.Install a GNSS receiver antenna on each side of the rotating carrier in opposite directions,the GNSS signal can be received stably under any calendar element.However,when the carrier rolls,the traditional kalman measurement results will hop due to GNSS error.In this paper,a weighted Kalman filtering algorithm is proposed for a user model with low-speed rotation as an example.In applying this algorithm,the ratio between the residual of the observed quantity and the mean square of the measurement is used to adjust the variance of the overrun observation and reduce its filtering gain to weaken the weight of the overrun observation,so that all visible satellites can participate in the filtering and thus the positioning results will not hop.Semi-physical simulation further verifies that compared with traditional Kalman algorithm,the usability of the positioning results(3D positioning error less than 30 m)of the new algorithm is improved from 89%to 99%,and the 3D velocimetric error(1σ)is improved from 0.19 m/s to 0.11 m/s.