针对舰载条件的捷联惯导粗对准问题,提出了一种简单可行的最优粗对准方法。根据双矢量定姿的原理,分别将两个观测矢量之一作为基准,通过两次三轴姿态测定算法得到两个姿态矩阵,然后根据观测矢量的方差特性加权得到精度最优的姿态阵。阐述了三轴姿态测定算法的基本原理,分析了最优三轴姿态测定算法与基于高斯马尔科夫估计的三轴姿态测定算法的统一性,解析了基于最优三轴姿态测定算法的舰载惯导系统粗对准方案,并对传统三轴姿态测定算法和最优三轴姿态测定算法进行了应用比较。蒙特卡洛50个样本的仿真结果表明,采用最优三轴姿态测定算法明显优于传统三轴姿态测定算法,可使得东向、北向和天向姿态误差角均值分别为4.78,9.21和0.29,标准差分别为0.11,0.07和1.08,水平失准角最大值9.37,方位失准角最大值2.8,能够有效确定出载体的粗略姿态,在此基础上能更好实现该状态下的舰载惯导精对准。 Aiming at the problem of rough alignment of SINS with shipborne condition, a simple and feasible method of rough alignment is proposed. According to the principle of bi-vectorial attitude, two attitude matrices are obtained by using two observation vectors respectively. Two attitude matrices are obtained by two-way three-axis attitude determination algorithm, and then the attitude matrix with the best accuracy is weighted according to the variance characteristics of the observation vectors. The basic principle of the three-axis attitude determination algorithm is described. The unification of the optimal three-axis attitude determination algorithm and the three-axis attitude determination algorithm based on Gaussian Markov estimation is analyzed. Based on the optimal three-axis attitude determination algorithm, Rough alignment scheme of inertial navigation system, and the application of the traditional three-axis attitude determination algorithm and the optimal three-axis attitude determination algorithm are compared. The simulation results of 50 Monte Carlo samples show that the optimal three-axis attitude measurement algorithm is significantly superior to the traditional three-axis attitude measurement algorithm, so that the average value of the attitude error angles in the east, north and sky direction are 4.78, 9.21 and 0.29 respectively, The standard deviations are 0.11, 0.07 and 1.08 respectively, the maximum horizontal misalignment angle is 9.37 and the maximum azimuth misalignment angle is 2.8, which can effectively determine the rough attitude of the carrier. Based on this, the shipboard habit Guide precision alignment.