| 摘要: |
| 蓄洪排沙足冯家山水库的排沙方式．然而在实际调度运行过程中，对入库洪世和出库排沙泄量的调配缺乏共性关联，排沙泄世未能从定性定量上予以科学界定。针对此间题，文章对冯家山水库蓄洪排沙过程进行了研究，采用RBF网络建立了该水库出库含沙量预测模剐，模型根据沙峰、洪峰入库时间与开闸排沙时间的不同分别选择网络结构。采用冯家山水库历史排沙资料对模型进行检验的结果表明，模型训练及检验结果确定性系数均较大。可见，采用RBF网络建立的山库含沙量预测模型是可行的。 |
| 关键词: RBF网络 冯家山水库 出库含沙量 预测 |
| DOI： |
| 分类号: |
| 基金项目:高等学校优秀青年教师教学科研奖励计划项目（2001-282） |
|
| The application of RBF neural network in forecast of out flow sediment concentration of Fengjiashan Reservoir |
|
|
| Abstract: |
| Storing flood and discharging sediment is the most effective discharging sediment measure of Fengjiashan reservoir.But in fact,allocation between input flood hydrograph and outflow sediment concentration hydrograph lacks general relevancy,and discharging sediment hydrograph has not been defined scientifically in quality and quantity.So output sediment concentration hydrograph of Fengjiashan reservoir is researched.RBF ANN is used to establish forecast model of the outflow sediment concentration of the reservoir.The two kinds of ANN architectures are selected to establish the forecast model based on the difference between the time of entering the reservoir of the sediment peak and flood peak and the time of opening up gate for discharge.At the same time,previous data of discharging sediment of the reservoir are used to check the model,not only DC of the training but also those of the testing are bigger.This shows that it is feasible that RBF ANN is used to establish the forecast model of outflow sediment concentration. |
| Key words: RBF neural network Fengjiashan Reservoir out flow sediment concentration forecast |
