SUKF滤波在SINS大失准角初始对准中的应用研究

Application of SUKF filtering in initial alignment of large misalignment angle of SINS

针对捷联惯导系统在大失准角初始对准过程中的非线性滤波问题,提出了基于Scaled UT变换的Scaled-Unscented卡尔曼滤波(SUKF),并将SUKF应用到基于四元数误差的非线性对准模型中。在仿真中对比了SUKF和扩展卡尔曼滤波(EKF),结果表明,在大失准角的情形下,SUKF比EKF对准速度更快,估计精度更高。

The system error equation of Strap-down Inertial Navigation System （SINS） for large misalignment angle is nonlinear, thus it is urgent to solve the problem of initial alignment in SINS when the error equation is nonlinear. An approach of Sealed-Unscented Kalman Filter（SUKF） based on Scaled-Unscented Transformation （SUT） is presented. SUKF is then applied in the nonlinear alignment model based on the error of quatemion. The initial alignment of large misalignment angles is simulated by using Extended Kalman Filter（EKF） and SUKF respectively. The results indicate that SUKF is better than EKF in convergence speed and estimation precision.