基于混合迭代UKF的捷联惯导对准算法设计

捷联惯导系统陀螺噪声过大时,自对准无法使方位失准角估计收敛到真值,设计了一种磁强计辅助对准方法,满足对准精度要求并可同步估计磁场强度和磁倾角。非线性滤波方法UKF相比EKF估计精度高但收敛速度慢,迭代UKF算法可提高收敛速度但计算量也显著增大。将超球面分布采样点变换（SSUT）应用于UKF迭代过程,设计了一种混合迭代UKF算法。仿真表明该混合迭代UKF算法加快了收敛速度,保证了滤波精度,同时又减少了计算量。

Abstract：

The azimuth misalignment-angle estimation of SINS could not converge to the true value without assistance duo to the over too strong noise of gyros.An alignment method assisted by magnetometers was designed to estimate the misalignment angles as well as the magnetic field intensity and the magnetic obliquity.Nonlinear filtering method UKF was more precise than EKF but converged slowly.Iterated UKF could promote the convergence rate but increased the calculation amount greatly.A combined iterated UKF algorithm was proposed by applying spherical simplex unscented transform to the iterated process.Simulation indicates that the algorithm promotes the convergence rate,guarantees the precision,and reduces the calculation amount.

Algorithm Based on Combined Iterated UKF for SINS Alignment

The azimuth misalignment-angle estimation of SINS could not converge to the true value without assistance duo to the over too strong noise of gyros.An alignment method assisted by magnetometers was designed to estimate the misalignment angles as well as the magnetic field intensity and the magnetic obliquity.Nonlinear filtering method UKF was more precise than EKF but converged slowly.Iterated UKF could promote the convergence rate but increased the calculation amount greatly.A combined iterated UKF algorithm was proposed by applying spherical simplex unscented transform to the iterated process.Simulation indicates that the algorithm promotes the convergence rate,guarantees the precision,and reduces the calculation amount.