The initial alignment provides an accurate attitude for SINS (strapdown inertial navigation system). By further estimating the IMU's bias and misalignment angle, the recursive Bayesian filter is accurate. However, the prior heading error has significant influence on the convergence speed and accuracy. In addition, the accuracy will be limited by its iteration at a single time-step. Coarse alignment method OBA (optimization-based alignment) uses MLE (maximum likelihood estimation) to find the optimal attitude quickly. However, few methods consider the IMU bias and misalignment angle, which will reduce the attitude accuracy. In this paper, a unified method based on FGO (Factor graph optimization) and IBF (inertial base frame) is proposed. The attitude is estimated by MLE, IMU bias and misalignment angle are estimated by MAP estimation. The state of all time steps is optimized together to further improve the accuracy. Physical experiments on the rotation MEMS SINS show that the heading accuracy of this method is improved in limited alignment time.

翻译：初始对准为SINS（捷联惯性导航系统）提供精确的姿态。通过进一步估计IMU的零偏和失准角，递归贝叶斯滤波器能够达到较高精度。然而，先验航向误差对收敛速度和精度有显著影响。此外，单时间步迭代会限制其精度。粗对准方法OBA（基于优化的对准）利用MLE（极大似然估计）快速求解最优姿态，但鲜有方法考虑IMU零偏和失准角，这会降低姿态精度。本文提出一种基于FGO（因子图优化）和IBF（惯性基坐标系）的统一方法：姿态通过MLE估计，IMU零偏和失准角通过MAP估计求解，所有时间步的状态被联合优化以进一步提升精度。旋转MEMS SINS实物实验表明，该方法在有限对准时间内提升了航向精度。