Optimization for four-sample rotation vector attitude estimation algorithm

The attitude estimation algorithm is one of the key technologies for precision navigation of strap-down inertial navigation system (SINS). In this paper, a high-precision attitude estimation algorithm is proposed to update the attitude for SINS. Specifically, the proposed algorithm, improved four-sample of double-loop algorithm, (hereinafter referred to as IFSDL) is based on the four-sample rotation vector algorithm and utilizes the double-loop iterative approach. IFSDL makes it possible to improve precision without increasing computational complexity. The advantage ensures it to be competent for the attitude estimation in the case of high maneuver. Under the classical coning motion, this paper analyzes and compares the attitude error of IFSDL with that of conventional four-sample algorithm. Additionally, the drifts reduction ability of IFSDL is verified through theoretical analysis and simulation experiment.