Fast Linear Attitude Estimation and Angular Rate Generation

This paper focuses on the approach to attitude estimation and virtual gyroscope compensation. First, the rotation transformation is built between the body frame and the reference frame by means of an accelerometer-magnetometer triad. Fast linear attitude estimation method is used in the process to improve the computation efficiency. The attitude quaternion is invoked for parameterization of orientation in Kalman filtering. Furthermore, to generate the virtual-gyro output in the case of gyroscope failures, virtual-gyro Kalman filter is established for angular rate estimation. Experiments are carried out to illustrate the validity and efficiency of the proposed attitude and angular rate estimation approaches.