基于平移不变量的激光陀螺信号快速滤波方法  

A Fast Denoising Algorithm Based on Translation Invariant for Laser Gyro Signal

作　　者：彭云辉[1] 刘云峰[1] 杨小冈[1] 缪栋[1] 

机构地区：[1]第二炮兵工程学院,陕西西安710025

出　　处：《上海航天》2007年第2期57-60,共4页Aerospace Shanghai

摘　　要：为提高激光陀螺的测量精度,提出了一种平移不变量小波快速滤波方法。对信号作循环平移,进行软(硬)阈值小波消噪处理再作逆平移,通过多次“平移——滤波——逆平移”过程,最后平均所获结果,有效消除软阈值小波滤波中的伪吉布斯现象。讨论了算法中平移量对滤波效果和计算复杂度的影响。用Allan方差法对一组激光陀螺实测零漂信号的不同方法滤波效果分析结果表明,该滤波法能在不影响滤波效果的前提下较大幅度减少计算量。A fast denoising method based on translation invariant of wavelet filtering was put forward to improve the measuring precision of laser gyro in this paper. The procedure of the algorithm was that the signal was translated at first, then the data was translated in reverse after the shifted signal was denoised by the soft or hard threshold, and last the averaged results were obtained after many cycles, which would weaken the pseudo-Gibbs phenomena in wavelet filtering within the soft threshold. The effect of the translation on filtering and computation complexity was discussed. The results for a set of zero drift signal of the laser gyro by various filtering methods using Allan variance showed that this method could shorten the computation time remarkably without influence on denoising.

关 键 词：激光陀螺 平移不变 滤波 平移量 ALLAN方差 伪吉布斯 消噪 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]