【摘要】 灌溉用水量的预测对灌区的灌溉管理起着重要的作用。运用神经网络中Elman反馈型神经网络建立了灌溉用水量预测模型,模型输入层神经元数目为4,输出层神经元数目为1,隐含层神经元数目确定采用试验法,最终确定为10。预测结果表明:该方法与传统的预测方法相比,具有网络稳定性高,训练误差曲线比较平滑,模型预报精度较高等优点。更多还原

【Abstract】 The forecast of irrigation water volume plays an important role in the irrigation management.This paper used Elman neural network to establish the model of irrigation water use prediction,having four importing units and exporting units.The determination of concealed level neural single number used the cut-and-try method,ten units were used.The forecast result showed that the model had high network stabilization,the training error curve was smoother and it had high forecasting precision compared with the traditional forecast method.更多还原