摘要
用 Allan 方差法对激光陀螺漂移数据的各种误差源及其整个噪声统计特性进行细致的表征和辨识。根据有效信号与各种噪声信号的频谱差异,用 Mallat 算法对 RLG 漂移信号进行塔式分解,采用软阈值消噪法去除噪声。对小波去噪前后的激光陀螺零漂信号用 Allan 方差进行对比分析表明,小波变换有效的减小了激光陀螺零漂信号中的角度随机游走、角速度随机游走、速率斜坡和量化嗓声,提高了激光陀螺零偏输出的稳定性。
Variant kinds of ring laser gyro(RLG)noise and the statistical characters of general noise are identified and expressed by Allan Variance.According to the spectrum difference between effective signal and noise,the drift data of RLG is made'cone decomposition'by Mallat's Algorithm,and denoised by soft-threshold denoising method.The RLG drift data before and after Wavelet De-noise are analysed and compared by Allan variance,the result indicates that Wavelet Transform can effectively reduce the angle random walk,rate random walk,rate ramp and quantization noise,improve the bias instability of RLG drift data.
出处
《弹箭与制导学报》
CSCD
北大核心
2005年第S6期151-154,共4页
Journal of Projectiles,Rockets,Missiles and Guidance