SINS Rapid In-Motion Alignment Based on Nonlinear Filter

Zhilan Xiong, Yanling Hao, and Feng Sun

Abstract:	Strapdown inertial navigation system (SINS) is a dead-reckoning navigation system. Hence, the navigation performance of SINS heavily depends on the proper initial alignment, which determines the transformation matrix from the body frame to the navigation frame. The velocity and attitude matching scheme is known as one of the best combined matching algorithm of SINS in-motion alignment, since it can achieve submilliradian alignment accuracy without the lateral acceleration requirement and lengthy time requirement. In this paper, a new SINS/GPS in-motion alignment model based on the quaternion is presented for the velocity and attitude matching algorithm. It directly uses the quaternion form of the misalignment angles as the state variables. By the simulation of the on-sea alignment, six nonlinear filters are compared. These filters include the unscented Kalman filter (UKF), the unscented particle filter (UPF), the Gaussian particle filter (GPF), the Rao-Blackwellized unscented Kalman filter (RBUKF), the Rao-Blackwellized unscented particle filter (RBUPF) and the Rao-Blackwellized Gaussian particle filter (RBGPF). The simulation results show that the estimation value of the azimuth misalignment can converge at the round of the true value with the same convergent velocity as the level misalignment if using the proposed model in this paper. Although only 20 particles are used in each particle filter, the estimation values of the misalignments can rapidly converge at the round of the true values in less than 0.1 second (two steps in this paper). Moreover, the estimation accuracy of the azimuth misalignment could achieve sub-milliradian in less than 1 second for the small initial misalignments, subdecimilliradian in less than 20 seconds for the large initial misalignments. Furthermore, the estimation performance of RBGPF is best, and one of UPF is worst in these considered filters.
Published in:	Proceedings of IEEE/ION PLANS 2006 April 25 - 27, 2006 Loews Coronado Resort Hotel San Diego, CA
Pages:	86 - 93
Cite this article:	Xiong, Zhilan, Hao, Yanling, Sun, Feng, "SINS Rapid In-Motion Alignment Based on Nonlinear Filter," Proceedings of IEEE/ION PLANS 2006, San Diego, CA, April 2006, pp. 86-93. https://doi.org/10.1109/PLANS.2006.1650591
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In