基于逆向Kalman滤波的行进间对准

In-motion alignment scheme based on reverse Kalman filter

为实现车载捷联惯导行进间对准的快速性和高精度,提出一种基于逆向Kalman滤波算法的行进间对准方案。与传统的先粗对准再精对准的两段式对准法不同,所提出的方案将整个对准过程中采集存储的数据同时用于粗对准和精对准。在对准结束前使用逆向Kalman滤波对系统状态进行回溯估算,利用初始时刻的已知信息对系统进行补偿校正后,再正向组合导航至对准结束。通过对整个过程中采样数据的分时复用,达到无需静态粗对准,也无需再次位置校正的目的,降低了对车辆运动条件的约束,显著提高了车辆的机动性。跑车试验表明:该方案可以在10 min内完成初始对准,并达到较高的定向和定位精度,航向误差为0.042°(1σ),定位误差为5.99 m(RMS)。

A new scheme based on the reverse Kalman filter is proposed to achieve the rapidity and high precision of the in-motion alignment of the vehicle-mounted strapdown inertial navigation system.Different from the traditional two-stage alignment method,the proposed scheme uses the collected data for both coarse alignment and fine alignment.Before the end of alignment,the reverse Kalman filter is used to retrograde estimate the system state,the known information at the initial time is used to compensate and correct the system,and then the forward combined navigation is carried out to the end of alignment.Through the time-division multiplexing of the sampling data in the whole process,the purpose of eliminating the need for static coarse alignment and position-known point is achieved,which reduces the constraints on the carrier movement conditions and significantly improves the mobility of the carrier.The results show that the scheme can complete the initial alignment in 10 minutes,and achieve high direction and position accuracy.The heading error is 0.042°(1σ)and the position error is 5.99 m(RMS).