摘要
针对光纤陀螺(FOG)输出信号的随机噪声问题,提出了一种解耦自适应Kalman滤波方法用于FOG信号降噪。方法采用Allan方差分析法估计量测噪声的方差参数,Kalman滤波过程与量测噪声方差的估计过程完全独立,避免了Kalman滤波器与量测噪声估值器间的相互耦合。利用时间序列分析法对FOG的随机噪声进行建模,并在建立的二阶自回归模型的基础上,采用本文滤波方法对采集的FOG实测数据随机噪声进行了试验验证。与传统方法相比,提出的滤波方法具有更好的降噪效果,噪声均方误差降低40%以上。
To solve the problem of random noises in the output data of fiber optic gyroscope(FOG), a decoupling adaptive Kalman filter is proposed for FOG signal de-noising. The parameter of the measurement noise variance is estimated by Allan variance analysis method. The processes of measurement noise variance estimation and Kalman filter are independent of each other. Therefore, the coupling between Kalman filter and estimator of measurement noises is avoided. The model of FOG random noise is established by using time series analysis method. Then, based on the second-order regression model, the collected random noise of FOG measured data is tested by using the proposed filtering method. Compared with traditional method, the proposed filtering method has better de-noising effect, and the mean square error of noise can be reduced by at least 40%.
出处
《中国惯性技术学报》
EI
CSCD
北大核心
2014年第2期260-264,共5页
Journal of Chinese Inertial Technology
基金
国家自然科学基金(61374206
61104184)
国家重大科学仪器开发专项(2011YQ12004502)
海军工程大学自然科学基金(HGDQNJJ12028)资助
