预测日径流过程的双线性模型

A Genetic Bilinear Time Series Model for Predicting the Daily Flow Process

提出了建立双线性模型 ( BM)的一套简便实用的方案 .实例计算说明 :1这套方案在日径流过程预测中是可行而有效的 ;2通过利用预测过程中产生的残差信息进行反馈矫正 ,保证了 BM模型高的拟合精度和稳健的预测性能 。

A simple and practical scheme is presented for developing a bilinear time series model (BM). The computational results of an example show that the scheme applied to predicting the daily flow process is feasible and efficient and that BM can ensure high fitting precision, robust forecasting and good adaptability to complex nonlinear dynamic systems by using the feedback information of prediction residual errors. As a general approach, the scheme is of major theoretical value and wide ranging applicability for predicting nonlinear time series.