摘要
小波具有多分辨率分析特性,利用小波阈值去噪可以有效地处理MEMS陀螺信号的噪声。针对常用阈值小波去噪方法的不足,提出了一种改进的小波阈值方法。详细介绍了该方法的原理与实现过程,并将其应用到MEMS陀螺信号处理过程中。使用db4小波,应用不同的阈值函数对MEMS陀螺信号进行4尺度分解。仿真实验分析比较表明,改进的小波阈值方法可以有效地剔除信号中的噪声,抑制了MEMS陀螺仪的随机漂移;与现有的阈值方法相比,信噪比提高了18.3%。
The wavelet has the multi-resolution analysis characteristics, and the wavelet threshold denoising method can be applied to process the noise of MEMS gyro signal. Based on the analysis to the existing threshold wavelet de-noising methods, an improved wavelet threshold de-noising method is presented. The principle and implementation of this method are presented in detail. The method was used to process the MEMS gyro＇s signal. By using db4, different threshold functions were adopted for signal decomposition of the gyro. Experiment analysis and comparison indicated that the proposed method can effectively eliminate the noise, and restrain the random drift of MEMS gyro. Compared with traditional threshold method, the SNR is increased by 18.3%.
出处
《电光与控制》
北大核心
2009年第12期61-64,共4页
Electronics Optics & Control
基金
国家自然科学基金项目(60304004)
关键词
MEMS陀螺
小波分析
阈值去噪
阈值函数
随机漂移
MEMS gyro
wavelet analysis
threshold de-noising
threshold function
random drift
