基于改进小波阈值的MEMS陀螺去噪方法

MEMS陀螺由于结构不完善而存在较大的随机误差,为了推进MEMS陀螺的实用化水平、提高测量精度,首先对MEMS陀螺进行了Allan方差分析,得到了MEMS陀螺随机误差的主要成分;然后通过分析软阈值函数和硬阈值函数在直线上的降噪结果,得到了两种阈值函数的缺陷,为了解决软阈值和硬阈值函数的问题,推导了改进的小波阈值函数。通过对比各种阈值函数和阈值量化准则的降噪效果,确定了最终的降噪方案,将此方案应用在陀螺的动静态输出中,使陀螺输出数据的稳定性提高了一个数量级,数据的平均值也更加接近真实值。

MEMS Gyroscope Denoising Method Based on Improved Wavelet Threshold

There existing great random error in MEMS gyro output due to its structure defects,to increase measurement accuracy and advance practical level,firstly this essay analyzes MEMS gyro output through Allan Variance approach,and the main component of random error is gotten. Then through analyzing denoising result of hard threshold and soft threshold,shortcomings of these two thresholds are found. To solve the shortcomings of hard threshold and soft threshold, improved threshold function is constructed. Comparing denoising effect of different threshold function and threshold quantitative guidelines,the eventual denoising project is confirmed. Using this project in dynamic and static output of gyro,the result indicates that stability of gyro output increases a degree of magnitude,and the mean is closer to the real value.