论文部分内容阅读
针对光纤陀螺输出是非平稳信号的特点,提出了小波-ARMA模型建模方法 ,并与ARIMA模型进行了对比分析;针对滤波方程中的量测噪声不能确定的问题,提出了基于量测噪声自适应的滤波方法,并与卡尔曼滤波进行了对比分析。研究结果表明:滤波方法相同时,小波-ARMA模型的滤波效果优于ARIMA模型;模型相同时,自适应滤波效果优于卡尔曼滤波;且建模精度大于滤波方法对滤波效果的影响。Allan方差分析表明:基于小波-ARMA模型的自适应滤波能很好地滤除光纤陀螺中的量化噪声,提高陀螺的输出精度。
Aiming at the characteristics that the output of FOG is non-stationary, a wavelet-ARMA model modeling method is proposed and compared with the ARIMA model. Aiming at the problem that the measurement noise can not be determined in the filter equation, The filtering method is compared with Kalman filtering. The results show that the filtering performance of the wavelet-ARMA model is better than that of the ARIMA model when the filtering methods are the same; the adaptive filtering effect is better than the Kalman filtering when the model is the same; and the modeling accuracy is greater than the filtering effect on the filtering effect. Allan variance analysis shows that the adaptive filtering based on the wavelet-ARMA model can well filter out the quantization noise and improve the output precision of the gyro.