本文選題：隨機(jī)誤差　切入點(diǎn)：自回歸滑動平均(ARMA)模型　出處：《壓電與聲光》2017年01期 　論文類型：期刊論文

【摘要】：針對半球諧振陀螺(HRG)隨機(jī)誤差影響慣性測量單元測量精度的問題,提出了一種改進(jìn)的基于自回歸滑動平均(ARMA)模型和自適應(yīng)濾波(AKF)的隨機(jī)誤差處理方法。該文對預(yù)處理的數(shù)據(jù)進(jìn)行了自相關(guān)和偏相關(guān)特性分析,判斷隨機(jī)誤差的適用模型,以及利用貝葉斯信息準(zhǔn)則(BIC)準(zhǔn)則估計(jì)ARMA模型的階數(shù),通過長自回歸模型計(jì)算殘差法獲取模型參數(shù),引入加權(quán)自適應(yīng)因子在線調(diào)整一步預(yù)測誤差陣和量測噪聲矩陣用于改進(jìn)濾波方程,并比較了5項(xiàng)主要誤差系數(shù)值。結(jié)果表明,改進(jìn)的算法能夠有效抑制隨機(jī)誤差,為HRG的隨機(jī)誤差建模補(bǔ)償提供了新方法。
[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.
【作者單位】： 火箭軍工程大學(xué)控制工程系;
【基金】：國家自然科學(xué)基金資助項(xiàng)目(61174030)
【分類號】：TN96

【參考文獻(xiàn)】

相關(guān)期刊論文 前9條

1 林青;戴慧s，

本文編號：1582260