Abstract:In order to improve the accuracy of MEMS gyro and to meet wide application requirements, an automatic regressive model (AR model) and output model of single MEMS gyro are established. Based on the two models, the state space equation and measurement equation of virtual gyroscope are found. A Kalman filter to combine outputs of gyro array is designed, and to enhance real-time ability, a novel Kalman filter is also constructed. Constant-rate and angular vibration experiments are carried out to verify the validity of the data-combining approach. The results indicate that relative to single gyro, the virtual gyro’s accuracy can increase 18.18 times in the constant-rate experiment and 5.36 times in the angular vibration experiment.