In this paper, the issue of fine alignment method about Doppler-assisted Strapdown inertial navigation system is studied. The main errors of the typical fine alignment method are analysed. The problem of the typical fine alignment method is the speeds measured by Doppler velocity log have to be transformed from the body frame to the navigation frame. But the relationship between the body frame and the navigation frame have not be established exactly as the fine alignment is not completed. This brings large errors for the fine alignment. Then, an improved Doppler-assisted alignment method is proposed. The Kalman filter dynamic model in the body frame for Dopplerassisted SINS is deduced. Because the DVL speeds are used directly, it avoids the problem that the typical fine alignment method exists. Simulations show that the improved Dopplerassisted alignment method obtains higher alignment precision. As open-loop Kalman filter is used, the new method has a high project value.

Regarding the problem of the Autonomous Underwater Vehicle (AUV) departing from the mother ship and starting up, this paper proposed a divided difference filter (DDF) basing on the Stirling interpolation formula. When AUV separated from the mother ship and restarted, the coarse alignment was not ideal. Its level misalignment error precision was high, but its heading misalignment error precision was low. The large heading misalignment error model was applied to the fine alignment on moving base. The large heading misalignment error model was nonlinear. The first order divided difference (DD1) filter and the second order divided difference (DD2) filter were adopted to the nonlinear filtering. Simulation results show that the DD1 alignment precision is equivalent with Extended Kalman Filter (EKF), but the alignment time shorts by 33%. The DD2 alignment precision is equivalent with Unscented Kalman Filter (UKF), but the alignment time shorts by 28%. The alignment precision raises by 60% compared to the EKF.Index Terms -Autonomous Underwater Vehicle; Divided Difference Filter; large heading misalignment error; nonlinear filtering.

No abstract

No abstract

In order improve the performance of Strapdown Inertial Navigation System (SINS) alignment in moorage, a novel alignment method is proposed. With the aid of a cascade low-pass FIR filter, the gravity vector is separated from the measurements of accelerometers due to the high frequency characteristic of the disturbance accelerations and the low frequency characteristic of the gravity vector. Two non collinear velocities both in the inertial frame and in the concretionary are integrated by the gravity vector and the cosin matrix between the two frames is calculated. With the other matrix which can be calculated easily, the alignment is accomplished. Results of experiment show that the accuracy of the alignment method proposed in this paper converges much faster than the traditional alignment methods. However, the new method is a open-loop method, thus the precision is not significantly improved.