Redundant Rotational Inertial Navigation System Attitude Information Fusion Method Based on Asynchronous Rotation Modulation

Rotational inertial navigation systems (RINS) are widely used for their ability to autonomously suppress long-term errors. However, the rotational modulation process inevitably introduces short-term disturbances. These disturbances primarily stem from non-orthogonal errors and angular measurement inaccuracies, significantly degrading attitude stability and directly impacting performance in high-precision measurement scenarios and rapid alignment of sub-inertial systems. To address this issue, we propose a novel attitude information fusion method for redundant RINS. First, an asynchronous rotation modulation strategy is designed. By staggering the rotation sequences of the two RINSs, we ensure that at least one system remains stationary at any given time without requiring hardware modifications or alterations to core navigation algorithms. Theoretical analysis reveals that direct attitude switching under this scheme introduces step errors due to residual installation errors. To overcome this limitation, a fusion algorithm based on relative attitude matrix transmission is developed. Using the pre-rotated attitude of a single system as the reference, it continuously integrates the stable relative attitude changes of the stationary system during rotation, dynamically suppressing error spikes. Experimental results demonstrate that the fused attitude reduces maximum heading error by 31.90% and improves peak-to-peak error stability by at least 15.9%. This method effectively enhances the attitude output quality of redundant RINS, enabling more reliable and stable navigation in maritime environments.