A New Initial Alignment Algorithm without Latitude For SINS On In-Motion Condition

The initial alignment accuracy greatly affects the navigation service performance of the strapdown inertial navigation system (SINS). The local latitude information obtained from the global satellite navigation system must be provided to assist SINS in traditional initial alignment. However, in some complex environments such as tunnels and seabed, the traditional initial alignment methods will be unavailable because the satellite signals losses of signal lock. To achieve initial alignment without latitude on in-motion condition, an initial alignment method based on outlier robust Kalman filter is proposed. The measurement noise modeling method based on the student's t distribution is used to improve the alignment accuracy and speed. The real SINS data collected from ships sailing in the ocean are utilized to test the proposed method. The experimental results have shown that when the local latitude information is lost, the alignment accuracy of roll, pitch and yaw angle are improved by 13.6%, 2.3% and 25.7% respectively compared with the traditional method. It has been proven that the proposed method can be applied to initial alignment without latitude on in-motion condition for the SINS.