Research on GFSINS/Star-sensor Integrated Attitude Estimation Algorithm Based on UKF

摘要

In this research, an effective nine-accelerometerconfiguration is designed to form a gyroscope free strapdowninertial navigation system (GFSINS), and the absolute valuemethod is adopted to calculate angular velocity. This paperproposes an UKF attitude estimation algorithm for the starsensor aided GFSINS integrated navigation system. The experimental results show that, UKF algorithm avoids the higherorder truncation error in EKF, and improves the steady-stateprecision and convergence speed to some extent. Especially incomplex models, UKF is free from computational complexity ofthe Jacobi matrix, thus highlighting the rapidity and improvingthe computation efficiency of the system. Consequently, thisnavigation methodology could be competent for some highprecision and long-endurance navigation tasks.