摘要
提出了径流长期分级预报的人工神经网络方法，给出了多层前馈网络的联合梯度算法。用辽宁大伙房水库的汛期长系列资料进行实例研究，取得了初步令人满意的结果。
An artificial neural network method is proposed for hydrological forecasting, the hybrid gradient method for multilayer feedforward neural network developed.and this neural network model applied to runoff forecasting of Dahuofang reservoir with satisfactory results.
出处
《水电站设计》
1995年第1期1-6,共6页
Design of Hydroelectric Power Station
关键词
水文预报
径流
多层前馈网络
联合梯度法
hydrological foreeasting, neural network, BP Algorithm,conjugate gradient.
