Micromachined Gyroscope Precision "virtual" Implementation Research

With the excellent features of small volume, low weight ,low price, long life-span and easiness for volume-production, microgyros based on MEMS have strong potential to replace conventional gyros. However, current state-of-the-art micromachined gyroscopes can not compete with the established sensors in high-accuracy application areas such as guidance and inertial navigation. So it is significant to develop a new method to improve the performance of microgyros not only for the development of this technology but also for enforcing the influence of our country in this field.In this paper one method based on multi-sensor fusion was presented to improve the accuracy of the micromachined gyroscopes, which was named as "Virtual Gyro".The virtual gyro consisted of homogeneous multi-sensor fusion and heterogeneous multi-sensor fusion, which were respectively realized to obtain the optimal estimation of microgyros' output. Firstly, the cross correlation between homogeneous sensors(microgyros) was used to establish the Kalman filter system, and the relationship between the accuacy of virtual gyro and the cross correlation was studied. Secondly, heterogeneous sensors' signals were imported to accomplish the calibration of microgyros' random noise, improving the virtual gyro's performance. Finally, the filter hardware based on DSP(digital signal processor) was design to complete the filter calibration independently.The main contents in this thesis are described as follows:(1) The basic principle of virtual gyro, stochastic error model of microgyros and Allan variance modeling scheme are studied. The problem of how to apply the results extracted by Allan variance modeling scheme directly to stochastic error model is solved for the first time in China.(2) A virtual gyro method based on homogeneous multi-sensor fusion is proposed. Then several microgyros of the same kind form a gyro array whose drifts are compensated by specifically designed static filter and dynamic filter according to Kalman theory. These filters are implemented in Matlab/Simulink software. The experiments showed that three gyroscopes with bias drift of 67 degree per hour were able to be combined into one virtual gyroscope with drift of 0.255 degree per hour and 36 degree per hour through the presented static filter and...