MEMS传感器随机误差分析及处理

摘要

The precision of MEMS sensor limits its scope of application. In order to reduce random errors in MEMS sensors to improve its performance,it is great significance to improve the accuracy of MEMS sensor by analyzing and processing its random error. Firstly, the main stochastic error of the MEMS gyroscopes and accelerometers is analyzed by the Allan variance methods. Then from the perspective of time-frequency,the wavelet transform analysis could separate effectively zero random error, colored noise, white noise from sensor random error. Finally, the analysis results of the experimental data using the wavelet de-noise methods show that,the wavelet analysis has the advantage of separating and reducing the random error.%MEMS传感器的精度限制了其应用范围，为减小MEMS传感器随机误差的影响，提高其精度，对其随机误差进行分析和处理具有重要意义。首先采用Allan方差法分析了MEMS陀螺仪和加速度计的主要随机误差。然后基于Allan方差分析结果，发现从时频角度，采用小波变换分析传感器的随机误差，可以对零偏，有色噪声，白噪声进行了有效分离。最后通过对小波降噪方法的探讨，发现小波分析具有分离、减小随机误差的优势。