A Single-Sample Rotation Vector Attitude Algorithm with Quadratic Optimization under Angular Rate Input

Abstract

Attitude solution is a key technology in inertial navigation. The rotation vector algorithm can compensate for the non-commutability error of rotation. In order to solve the problem that the multi-sample rotation vector algorithm will decrease the attitude update frequency, this paper proposes a single-sample rotation vector attitude algorithm under angular rate input that uses the angular increment information of the current and previous time and the angular rate information of the current and previous N sampling times. On the basis of the traditional conic optimization method of error compensation coefficient, the algorithm coefficient optimization criterion of the periodic term is designed, and the error compensation coefficient of the periodic term is optimized twice. The experimental results show that the single-sample rotation vector attitude algorithm proposed in this paper with quadratic optimization under angular rate input not only has a high attitude solution accuracy, but also increases the attitude update frequency to the inertial navigation sampling frequency, that is, compared with the N sub-sample rotation vector algorithm, the attitude update frequency can be increased by N times, and has certain engineering application value.