| 摘要: |
| 针对目前神经网络应用于洪水预报时存在的不足,引入遗传算法、模糊神经网络对BP网络模型进行了改进,建立了基于改进BP网络模型的洪水预报模型,并将改进的BP网络模型应用于文峪河洪水过程的预报.预测结果表明,过程预报合格率达93.54%,达到了水文预报规范的要求;与传统BP网络相比,改进算法可提高洪峰的预报精度. |
| 关键词: 洪水预报 改进BP网络模型 遗传算法 模糊神经网络 |
| DOI： |
| 分类号: |
| 基金项目:国家“863”计划项目“感应式数字液位传感器及自动化临控系统项目”(2004AAA001050)子项目“洪水预报监控系统及水资源调度管理” |
|
| Research on flood forecast based on improved BP model |
|
|
| Abstract: |
| Aimed at the problem in ANN used for flood forecast ,this article introduced genetic algorithm,fuzzy algorithm to ameliorate the BP model,built the flood forecast model based on improved BP model.The result of model forecast indicated that the improved model used in Wenyuhe River flood forecast can reach standardized precision.The forecast qualification rate of process is 93.54%.Compared with classical model,the optimized model improved the forecast precision of flood peak greatly. |
| Key words: flood forecast improved BP modal genetic algorithms fuzzy ANN |
