基于地轴矢量解算的SINS无纬度支持自对准方法

针对无纬度条件下SINS初始对准问题，提出了一种基于地轴矢量解算的对准方法。在静止基座下，利用陀螺仪敏感地轴矢量，直接建立导航系轴向矢量在载体系的正交投影，解析式确定姿态矩阵；晃动基座条件下，根据惯性系重力矢量观测，构建以地轴矢量为旋转轴的两组等角旋转矢量序列，利用QUEST算法求解旋转四元数实现对地轴矢量的优化解算，并以此建立导航系轴向矢量的载体惯性系投影，最后利用陀螺跟踪载体系相对惯性系的变化，确定载体系相对导航系的姿态关系。静止基座解算实验表明直接解析式对准与传统的解析式对准精度相当，但解算效率得到了提高；晃动基座仿真与船载实验均表明基于四元数的地轴矢量优化解算在精度上要优于三矢量几何解算方法，引入低通滤波环节后，对准精度进一步提高。

An Initial Alignment Method of SINS without Latitude Based on Calculation of Earth Axis Vector

 A method of SINS initial alignment under the condition of no latitude based on calculation of Earth axis vector is proposed. The solution in this paper relies on the detection of the vector of the Earth rotational angular velocity which is called as the Earth axis vector on reference frame to define the navigation frame (n-frame) relative to the reference frame which is the board frame (b-frame) or board inertial frame (b0-frame).When the SINS under the static base, the orthonormal attitude matrix can be determined by building the orthonormal projection of the n-frame axis on the b-frame which is derived from the measurement of the Earth rotational angular velocity and the gravitational acceleration using gyroscopes and accelerators. On sway base, firstly, after building two sequences of the equivalent angle rotation around the Earth axis vector according to the measurement of gravity vector, then the Earth axis vector can be calculated using the solution of the QUEST algorithm. Thus, the b0-frame orthonormal projection of the n-frame axis can be obtained. Lastly, the relationship of the b-frame and n-frame can be determined according to the change of the b-frame relative to the b0-frame using the gyroscope data. The calculation test of the static base shows that the precision of the alignment method proposed is equivalent to the traditional method with the efficiency being improved. The sway-base simulation and ship-borne test show that the calculation precision based on the QUEST algorithm is superior to the method based on the geometry of three vectors, and then the precision can be further improved with the application of the low-pass filter algorithm.