A rapid and high-accuracy rotation alignment method based on bidirectional process for dual-axis rotational inertial navigation system

For the alignment in rotational inertial navigation system (RINS), which takes velocity errors of the navigation calculation as the filter measurements, the velocity error term caused by the coupling of the accelerometer's inner lever arms (ILAs) and the gimbals' angular motion will lead to a decline in the alignment accuracy. In view of that, a simplified model of the ILAs is proposed firstly in combination with the specific dual-axis rotation scheme, and then the ILAs are extended into state variables for filtering estimation, thereby improving the alignment accuracy. On this basis, in order to further improve the alignment rapidity, the bidirectional process-based rotation alignment method for dual-axis RINS is proposed in this paper. The bidirectional process, which combines the backward-forward RINS calculation and the backward-forward Kalman filtering, makes the best use of the stored IMU data repeatedly and is able to quickly obtain the IMU attitude matrix for RINS. The simulations and experiments show that the proposed bidirectional process-based rotation alignment method is able to accomplish the rapid and accurate alignment for the dual-axis RINS.