灰色自记忆神经网络模型在年径流预测中的应用

径流序列是一种非线性、弱相依高度复杂的动力系统,径流预测还处于探索阶段,提高径流预测精度关键在于对有限样本包含的信息进行充分发掘.采用一种理论和方法很难充分挖掘径流序列所包含的信息.基于组合预测思想,应用灰色理论,自记忆性原理与BP神经网络三种理论与建模方法对年径流序列进行挖掘,在此基础上提出年径流预测的灰色自记忆神经网络模型.结果表明,模型能够很好地反映径流的变化规律与极值趋势,具有较好的拟合与预报精度.

Application of cray-self-memory-neural network model to prediction of the annual runoff

Runoff time series is a non-linear,weakly dependent and complicated dynamic system.The prediction of it is still on the exploring stage.The key of improving the accuracy of runoff prediction is to dig the information included in the limited sample sufficiently.However,one theory or method alone can not work elfectively.On account of the thought of integrative prediction,three modelings are combined to forecast the annual runoff.It is shown that gray self-memory neural network model can reflect the changing law and extremum trend of runoff series sufficiently,it has higher prediction accuracy and may be fit for annual runoff prediction.