Novel nonlinear filter for sins initial alignment with large misalignment angles

In order to improve the precision and convergence rate of SINS initial alignment with large misalignment angles, a novel nonlinear filtering method, referred as square root simplex unscented quadrature Kalman filter (SR-SUQKF) is introduced into the initial alignment in this paper. By using the method of simplex unscented transformation to select a few cubature points and combining with Gauss-Laguerre quadrature rule to improve the numerical accuracy, a new set of high order sampling points with determined number and weighting coefficient are constructed and used for filtering with a square-root extension to improve the numerical stability. By means of mathematical simulation and turntable experiment, the filter performance of square root cubature Kalman filter (SR-CKF) and SR-SUQKF using three points in one dimension in SINS/GPS initial alignment on a moving base with large misalignment angles is analyzed and compared. The results of both experiments show that under the condition of large misalignment angles, the three-point SR-SUQKF can perform the initial alignment with high accuracy and high speed, the alignment accuracy and convergence rate of the three-point SR-SUQKF are much better than that of SR-CKF. The SR-SUQKF can be effectively used for the initial alignment with large misalignment angles in SINS/GPS and would be able to overcome inherent disadvantages associated with the conventional nonlinear filters.