水力发电2017，Vol.43Issue(8)：22-25,38,5.
基于反向传播算法的崇阳溪流域洪水流量预报
Research on Flood Flow Forecast of Chongyang River Watershed Based on Back Propagation Algorithm
摘要
Abstract
Based on the principle of Back Propagation algorithm,the number of hidden layer nodes of BP model is determined by trial and error method and the momentum item is added,the steepest descent momentum Back Propagation Neural Network forecast model is established for the flood flow process of Jianyang Hydrological Gauging Station in Chongyang River Watershed.The model uses corresponding flood flow processes of two hydrological gauging stations in the branches of Chongyang River,which are Masha Station in Masha River and Wuyishan Station in Chongyang River,and at the same time,the autoregressive flood flow of Jianyang Hydrological Station is also considered.The results show that the forecast accuracy of model meets application requirements.The forecast model could be used as one of flood flow prediction method and the scientific basis for flood control and flood preparation of watershed.关键词
洪水预报/流量/BP算法/神经网络Key words
flood forecasting/flow/Back Propagation algorithm/neural network分类
建筑与水利引用本文复制引用
金保明,李芸,张烜,高兰兰,周传争,付晓雷..基于反向传播算法的崇阳溪流域洪水流量预报[J].水力发电,2017,43(8):22-25,38,5.基金项目
福建省自然科学基金资助项目(2016J01734) （2016J01734）
河海大学水文水资源与水利工程科学国家重点实验室开放基金面上项目(2015490311) （2015490311）
