MA模型参数估计的两段最小二乘法及其在自校正跟踪滤波器中的应用

Two-Stage Least Squares Method of Parameter Estimation for MA Models and Its Application to a Self-tuning Tracking Filter

提出了滑动平均(MA)模型参数估计的两段最小二乘法。首先用递推最小二乘法对MA模型拟合一个高阶自回归(AR)模型,然后再用最小二乘法解一个不相容的超定线性方程组得到MA模型参数估值。一个应用于自校正跟踪滤波器的仿真例子说明了其有效性。

Two-stage least squares method of parameter estimation for mov-ing average(MA) models is presented. First, the MA model is fitted by ahigh order autoregressive (AR) model. Secondly, the MA model paremeterestimates are obtained by solving a inconsistent overdetermined set of Linearequations. A simulation example with application to a self-tuning tracking fil-ter shows its effectiveness.