基于LMD-WSVD的半球谐振陀螺混合去噪方法

A hybrid denoising method for hemispherical resonant gyroscope based on LMD-WSVD

为了减少半球谐振陀螺(hemispherical resonant gyroscope,HRG)输出噪声对导航精度的影响,提出了一种基于局部均值分解排列熵小波变换奇异值分解的混合去噪方法。首先,使用局部均值分解方法对HRG信号进行分解,然后排列熵将其划分为两类:低频分量和混合分量;之后将小波变换和奇异值分解级联以构成两级滤波器,对混合分量进行降噪处理,最后重构得到最终的信号。通过实验验证了该方法的有效性,实验结果表明,相比原始信号,所提出的方法有效地减少了HRG的输出噪声,提高了其测量精度,其中角度随机游走降低了99.9%,零偏稳定性降低了60.3%。

In order to reduce the impact of hemispherical resonant gyroscope(HRG)output noise on navigation accuracy,a hybrid denoising method based on local mean decomposition-permutation entropy-wavelet transform-singular value decomposition is proposed.Firstly,the signal of the HRG is decomposed using the local mean decomposition method;then these components are divided into two categories:low-frequency components and mixed components by the permutation entropy;lately,for the mixed components,the wavelet transform and singular value decomposition are cascaded to form a two-stage filter for noise reduction,and finally,components are reconstructed to obtain the final signal.The effectiveness of the proposed method is verified by the experiments.The results show that the proposed algorithm can effectively reduce the output noise in the HRG and improve its measurement accuracy,among which the angular random walk is reduced by 99.9%and the bias instability is reduced by 60.3%compared with the original signal.