带有色观测噪声的多传感器ARMA模型信息融合辨识

摘要

对带已知有色观测噪声的未知自回归滑动平均模型(ARMA)模型,提出了一种两段信息融合辨识方法:第一段用递推辅助变量(RIV)算法得到自回归(AR)参数的局部和融合一致估值,第二段用Gevers-Wouters算法和用伪逆求解线性方程组方法得到滑动平均(MA)参数和噪声方差的局部和融合一致估值.该方法可用于语言增强信号处理问题.一个仿真例子说明其有效性.%For the unknown autoregressive moving average (ARMA) model with known colored measurement noise, a two-stage information fusion identification method is presented: In the first stage, the local and fused estimates of the autoregressive (AR) paraments are obtained by the reeursive instrumental variable ( RIV), and in the second stage, the local and fused estimates of the moving average (MA) paraments and noise variance are obtained by the Gevers-Wouters algorithm and by solving linear equation by the pseudoinverse. These fused estimators have consistency. This method can be applied to signal processing with respect to speech enhancement. A simulation example shows its effectiveness.