论文部分内容阅读
门限模型(Threshold Model)是描述非线性系统的一种行之有效的数学模型。由于门限的划分,在不同区域内建模的可用数据就会减少,使模型参数的估计精度降低。本文提出的加权门限模型改变了一般门限模型把不同区域的参数截然隔离的做法,采用加权的方法将不同区域的参数估计值结合起来,从而提高了模型参数的估计精度,减少了模型的预报误差和残差。
Threshold Model (Threshold Model) is a proven mathematical model of nonlinear systems. Due to the division of the thresholds, the available data for modeling in different regions will be reduced, and the estimation accuracy of model parameters will be reduced. The weighted threshold model proposed in this paper changes the general threshold model to completely isolate the parameters of different regions. The weighted method is used to combine the parameter estimates of different regions, which improves the estimation accuracy of the model parameters and reduces the prediction error of the model And residuals.