基于ARMA-AKF的HRG随机误差建模分析

发布时间：2018-03-08 03:35

本文选题：随机误差　切入点：自回归滑动平均(ARMA)模型　出处：《压电与声光》2017年01期 　论文类型：期刊论文

【摘要】：针对半球谐振陀螺(HRG)随机误差影响惯性测量单元测量精度的问题,提出了一种改进的基于自回归滑动平均(ARMA)模型和自适应滤波(AKF)的随机误差处理方法。该文对预处理的数据进行了自相关和偏相关特性分析,判断随机误差的适用模型,以及利用贝叶斯信息准则(BIC)准则估计ARMA模型的阶数,通过长自回归模型计算残差法获取模型参数,引入加权自适应因子在线调整一步预测误差阵和量测噪声矩阵用于改进滤波方程,并比较了5项主要误差系数值。结果表明,改进的算法能够有效抑制随机误差,为HRG的随机误差建模补偿提供了新方法。
[Abstract]:Aiming at the problem that random error of hemispherical resonance gyroscope (HRG) affects the measurement accuracy of inertial measurement unit, An improved ARMA model based on autoregressive moving average (ARMA) model and adaptive filter (AKF) is proposed to deal with random errors. In this paper, the autocorrelation and partial correlation characteristics of preprocessed data are analyzed to judge the applicable model of random errors. The order of ARMA model is estimated by Bayesian Information Criterion (ARMA), and the model parameters are obtained by long autoregressive method. The weighted adaptive factor is introduced to adjust the one-step prediction error matrix and the measurement noise matrix to improve the filtering equation, and five main error coefficients are compared. The results show that the improved algorithm can suppress the random error effectively. It provides a new method for HRG stochastic error modeling compensation.
【作者单位】：火箭军工程大学控制工程系;
【基金】：国家自然科学基金资助项目(61174030)
【分类号】：TN96

【参考文献】