多维AR(p)与指数平滑组合预测算法的研究及应用

Combined Prediction Approach of Multivariate Auto-Regressive and Exponential Smoothing and It's Applitication

介绍了时间序列多维AR(p)模型和指数平滑模型的理论和方法,将这两种模型结合起来构成新的组合预测模型,并用于预测测井曲线.应用实例证明,多维AR(p)与指数平滑组合预测模型比单独用多维AR(p)预测模型有更高的预测精度.表明该组合预测模型是一种非常有效的预测新方法.

In this paper the theory and methodology of Multivariate Auto-regressive and Exponential Smoothing model are proposed. A new combined prediction model based on Multivariate Auto-regressive and Exponential Smoothing model, is established to predict the logging curves. And a case is given. The result showed that the degree of accuracy was higher by using Multivariate Auto-regressive and Exponential Smoothing model than by using Multivariate Auto-regressive. So this new Combined Prediction Approach of Multivariate Auto-regressive and Exponential Smoothing is effective.