多维ARMA模型的新息定理及预测

Innovation Theorem and Forecast for Multivariable ARMA Model

利用线性最小方差估计方法，推导出多维ＡＲＭＡ模型的新息定理，并依据该定理得出仅需ＡＲ参数的线性最小方差新息预测公式，这在计算上大为简化，同时又保证有一定的精度。最后给出了仿真实例。

In this paper, an innovation theorem for multivariable ARMA model is presented by the method of the linear minimum variance estimation. According to the theorem, the forecasting formula can be derived directly just based on AR parameters of the series, hence the solution of the forecasting problem can be simplified greatly. An example is given to demonstrate the valid of the theorem at last.