Gyro fault prediction algorithm based on UKF

Aimed at the gradient failure of the gyro drift increases, an algorithm based on estimating the angular rate according to the UKF and attitude kinematic equation for gyro fault prediction is presented in this paper. We use quaternion to describe the attitude kinematics equation, and the UKF filter model is created, which takes the satellite attitude angle and the gyro angular rate as the state variable, and the attitude angle is based on a sun sensor and an earth sensor as the observed variable. According to the residuals of the estimated angular rate and gyro measurement values, the gyro failure prediction method is presented. This method may avoid the error caused by kinetic equation, which may be limited by the spacecraft´s inertial and control moment and the shortage of the EKF and the PF. A simulation system is developed. The result shows that the algorithm can predict the gradient failure of the gyro drift increasing timely and accurately. The model is simple, easy to build, and has less calculation, and has good engineering practicability.