摘要
给出了用人工神经网络 (ANN)对三峡宜昌站的日径流预测模型建模的过程 ,对 ANN输入变量的选择和个数的确定以及隐藏层、输出层单元数的确定等关键技术问题进行了探讨。所建立的基于 ANN的预测模型可以进行提前 7d的日径流预测 ,预测结果令人满意。
This paper presents the process of an artificial neural network (ANN) based daily streamflow forecasting model for the Three Gorges Yichang station. The ANN structure, including the selection of input variables, the number of input variables, the size of hidden layers, and the recursive ANN model, is discussed. The ANN-based model established can forecast daily streamflow profiles seven days ahead of time. The forecast result is satisfactory
出处
《水电自动化与大坝监测》
2002年第4期65-67,共3页
HYDROPOWER AUTOMATION AND DAM MONITORING
基金
武汉市晨光计划资助项目 (2 0 0 0 5 0 0 40 2 8)
关键词
人工神经网络
径流
预测
递归神经网络
artificial neural network (ANN)
streamflow
forecast
recursive ANN (RANN)
