摘要
1/fγ类随机噪声是影响压电陀螺精度的主要因素之一,其中随机信号包括白噪声和分形噪声.由于分形噪声具有长期相关性和自仿射性,采用传统的低通滤波方法难以达到有效的滤波效果.本文利用分形噪声在小波变换域的特性采用小波变换域参数估计的方法获得噪声参数,通过小波白化的方法消除分形噪声的长期相关性和自仿射性,最后应用小波软阈值的滤波方法去噪.经过对某型号压电陀螺信号进行滤波实验,结果表明这种基于小波分析的滤波方法是有效的.
1/fγ Stochastic noises in piezoelectric gyroscope,which contain white noises and fractal noises,have bad effects on its measurement accuracy.Because of the long correlation and self-affinity of fractal noises,it is difficult to eliminate noises using low-pass filtering method.Utilizing special properties of fractal noises in wavelet transform domain,parameter estimation technology in wavelet transform domain is applied to obtain parameter of noises.And using wavelet transform method erases the special prope...
出处
《电子器件》
CAS
2007年第5期1587-1590,共4页
Chinese Journal of Electron Devices
关键词
小波变换
压电陀螺
1/fγ类噪声
软阈值滤波
wavelet transform
piezoelectric gyroscope
1/f~γ type noise
soft-threshold de-noising
