Design of a gyroscope with minimum error covariance

Abstract

A new approach is suggested to design a gyroscope using a Kalman filter error covariance matrix in order to suppress high-frequency disturbance of gyroscope input. This is possible because the Kalman filter error covariance is dependent on gyroscope parameters and noise statistics. With given noise statistics, an optimal gyroscope can be designed by assigning gyroscope parameters to minimize the estimation error covariance. An optimization algorithm and simulation results are presented to show that a minimum error covariance region exists. Practical parameter values might be selected in this region by considering physical limitations. Thus, a gyroscope can be efficiently designed by employing the analytical method.