Abstract:

A novel optimal attitude and attitude-rate estimation algorithm is proposed to address the problem of autonomous attitude determination using gravitational filed and geomagnetic field observation. The algorithm is in the Kalman filter framework and employs weighted statistical linear regression and adaptive filtering to finish the measurement update. It can synchronously provide attitude-rate information .The simulation results show that the algorithm can triple approximately the measurement precision of what the common method did.

Key words: navigation, attitude determination, Kalman filtering, weighted statistical linear regression, adaptive filtering