摘要
建立了基于神经网络的洪水预报模型.该模型根据历史洪水样本,将河道预测点上游参考点的水位、流量以及预测点的水位作为神经网络的输入,预测点的流量作为神经网络的输出.采用BP算法训练网络,并将训练成功的洪水预报神经网络模型分别按8,16,24 h等预见期对新胡洼闸、西坝口闸水位进行预报,取得了较高的预报精度,验证了模型的有效性.
A prediction model based on the ability of information treatment and identification for neural networks is proposed. The in- puts of model include the water level and the flux of reference point in the upstream of the predicted point, and water level of predicated point. The outputs include the flux in predication point. The BP algorithm is applied to train the network. Based on the neural network model trained successfully, the water levels of in Xinhuwa gate and Xibeikou gate are forecasted with 8 h, 16 h, and 24 h step length, respectively. The results achieve higher forecast accuracy, and this verifies the validity of the model.
出处
《华北水利水电学院学报》
2012年第5期4-6,共3页
North China Institute of Water Conservancy and Hydroelectric Power
关键词
神经网络
BP算法
洪水预报
neural networks
BP algorithms
flood forecasting
