光纤陀螺经过确定性误差补偿后,各种随机噪声成为其误差的主要来源。为了减小光纤陀螺的随机误差,研究基于小波包分析的滤波方法,分析了小波包滤波的原理以及滤波过程中阈值及阈值函数的选取问题,并进行了相应的改进。然后分别采用Allan方差法和功率谱密度(PSD)方法分析滤波前后光纤陀螺的随机误差特性,对滤波的效果进行理论上的分析。最后,选取光纤陀螺VG951的漂移数据进行仿真验证,结果表明,基于改进的小波包滤波方法能够有效地减小光纤陀螺的随机误差,从而提高测试的精度。

After deterministic error compensation,various random noises become the main error sources of fiber optic gyro (FOG).To reduce the random errors of FOG,the denoising method based on wavelet packet filtering is studied,including the principle of wavelet packet filtering and the selection of threshold and thresholding operator,which is modified.The Allan variance method and power spectrum density (PSD) are adopted to analyze the random error characteristics before and after filtering,and the theoretical analysis of filtering effect can be realized.At last,simulation is carried out based on the drift data of FOG VG951,it is proved that the proposed method can decrease the random error of FOG effectively.