基于BP人工神经网络的枯水径流预报方案研究  

Research on Low-Flow Forecasting Scheme Based on BP Artificial Neural Network

作　　者：缪益平[1] 邓俊[2] 

机构地区：[1]二滩水电开发有限责任公司,四川成都610021 [2]中国水利水电科学研究院,北京100044

出　　处：《水文》2008年第3期33-36,13,共5页Journal of China Hydrology

摘　　要：介绍了BP人工神经网络的枯水径流预报方法,编制了锦屏一级水电站枯水径流预报方案。根据枯水径流预报方案的预报精度评定成果,总结了应用BP人工神经网络进行枯水径流预报的特点。研究表明基于BP人工神经网络的枯水径流预报方案能够满足水文情报预报规范,具有较好的实用性和可行性。This paper introduced the low-flow forecasting method of BP artificial neural network and made a low-flow forecasting scheme for Jinping-Ⅰ Hydropower Station. According to the precision evaluation of the low-flow forecasting scheme, the characteristics to apply BP artificial neural network in the low-flow forecasting was summarized. The study show that the low-flow forecasting scheme based on BP artificial neural network can satisfy the hydrological forecasting standard, with its good practicability and feasibility.

关 键 词：神经网络 水文预报 径流 枯水期 非线性 

分 类 号：P338.3[天文地球—水文科学]