摘要
光纤陀螺分形噪声往往具有长相关特性,导致陀螺输出信号慢漂移.为了滤除这类分形噪声,提出了一种分数阶预白化差分方法,通过该方法可以将分形噪声转化为高斯白噪声;利用小波变换,在小波变换域中采用Bayes软阈值去噪方法去除该白化噪声,从而形成了一种光纤陀螺分形噪声的预白化滤波方法.对VG949p型光纤陀螺实测数据处理结果表明,该方法能够有效地去除其中的分形噪声,抑制了陀螺低频漂移,同时也滤掉了其高频噪声项.
Fractal noise in fiber optic gyroscope（FOG） usually is long-term dependent,which causes FOG a rather slow drift.In order to filter this noise,a pre-whitening filter based on fractional difference（FD） was proposed.By this filter,fractal noise could be transformed into Gaussian white noise（GWN）.And through wavelet analysis,the pre-whitening noise could be eliminated by Bayes soft-threshold denoising in wavelet domain.On the two aforementioned steps,it came into a pre-whitening filter method of fractal noise.After the test data processing of FOG-VG949p,the results show that its fractal noise was removed efficiently,so the slow drift is restrained.The high frequency noise terms of FOG also greatly decrease.
出处
《北京航空航天大学学报》
EI
CAS
CSCD
北大核心
2011年第2期216-219,共4页
Journal of Beijing University of Aeronautics and Astronautics
基金
航空科学基金资助项目(20070851011)
关键词
光纤陀螺
随机误差
差分方程
小波分析
fiber optic gyroscope（FOG）
random errors
difference equations
wavelet analysis
