Fast Alignment and Calibration of Rotational Inertial System Based on Bilinear Embedding

In the system-level calibration of the rotational inertial navigation system (RINS), the effects of alignment and inertial measurement unit (IMU) output errors on inertial navigation results are nonlinear coupling. When the filtering algorithm corrects the attitude error and the output of the IMU in the form of feedback, the linearization error leads to a slower convergence rate and may even diverge the filter. This article proposed a novel method to avoid the convergence issue. The accurate linear error transfer equation is obtained using bilinear variable substitution, and the alignment and calibration problem is transformed into a three-degree-of-freedom unconstrained programming problem. This method does not need to make any small assumptions about the initial attitude deviation and has broad applicability. Simulation and experimental results show that the proposed method can quickly correct the RINS output under an unknown initial attitude angle. In the navigation experiment, the proposed algorithm's east and north velocity errors in 1 h are reduced by 40% and 97%, respectively, compared with the traditional algorithm.