硅微陀螺漂移数据滤波方法研究

Research on Filtering Method of Silicon Micro-Machined Gyroscope Drift

针对硅微陀螺测量信号特点,建立其测量信号一阶差分后的AR(2)信号模型,得到了硅微陀螺测量系统的状态方程和输出方程,根据该模型,采用鲁棒性很强的H-∞滤波方法,可以将静态漂移测量信号的标准差提高了一个数量级,动态信号的信噪比提高了6dB,大大提高了硅微陀螺的测量精度。H-∞滤波效果和实时性比小波变换要好。

Based on the measurement character of the silicon micro-machined gyroscope, an AR（2） model is built after the measurement signal is first order divided. The system function and the output function of the gyroscope are got. Depending on these functions, the superior robust H_∞ filtering is used to process the original gyroscope measurement. The filtering result shows that variance of the static drift can be improved more than ten times. Signal noise rate from the dynamics measurement can be improved 6 dB. H_ ∞ filtering is also better than the wavelet transformation in the filtering real-time and the filtering result.