Abstract:Aiming at the problems of low precision and poor adaptability of traditional moving base coarse alignment method for vehicle strapdown inertial navigation system (SINS), an optimal coarse alignment algorithm for moving base based on multi vector attitude determination is proposed. On the basis of the traditional initial alignment method based on gravity vector, the problem of solving attitude matrix is transformed into Wahba problem to make full use of the information of gravity vector at multiple times. The problem of solving Wahba is solved by SVD algorithm. Combining the output information of global navigation satellite system (GNSS) and SINS, the process equation and measurement equation are constructed. Sage-Husa adaptive filtering algorithm is used to solve the problem of measurement noise inaccuracy, and the transform matrix of carrier frame is continuously modified to obtain more accurate attitude conversion matrix. Simulation and semi physical experiments show that the improved algorithm can significantly improve the alignment accuracy, and the alignment azimuth error of turntable experiment is less than 0.05 °.