Design and analysis of a novel virtual gyroscope with multi-gyroscope and accelerometer array.

Abstract

A new virtual gyroscope with multi-gyroscope and accelerometer array (MGAA) is proposed in this article for improving the performance of angular rate measurement. Outputs of the virtual gyroscope are obtained by merging the signals from gyroscopes and accelerometers through a novel Kalman filter, which intentionally takes the consideration of the MEMS gyroscope error model and kinematics theory of rigid body. A typical configuration of the virtual gyroscope, consisting of four accelerometers and three gyroscopes mounted on designated positions, is initiated to verify the feasibility of the virtual gyroscope with MGAA. Static test and dynamic test are performed subsequently to evaluate its performance. The angular random walk (ARW) and bias instability, two static performance parameters of gyroscope, are reduced from 0.019°/√s and 14.4°/h to 0.0074°/√s and 8.7°/h, respectively. The average root mean square error (RMSE) is reduced from 0.274°/s to 0.133°/s under dynamic test. Compared with the published multi-gyroscope array method, the virtual gyroscope proposed here has a better performance both in static and dynamic tests, with improvement factors of ARW and RMSE about 44.1% and 44.5% higher, respectively.