基于BP神经网络和遗传算法的丰满水库洪水预报模型研究  

Flood forecast model of Fengman reservoir based on BP neural network and genetic algorithm

作　　者：李成林[1] 杨斌斌[1] 

机构地区：[1]水利部松辽水利委员会水文局,吉林长春130021

出　　处：《东北水利水电》2009年第8期57-58,62,共3页Water Resources & Hydropower of Northeast China

摘　　要：根据丰满水库的流域特性,尝试使用BP神经网络模型进行洪水预报。针对随机生成网络权重的盲目性,采用遗传算法进行初始权重优化。通过历史洪水检验,证实此模型在丰满水库的洪水预报中有很高的预报精度和应用价值。BP neural network model is tried to be used to forecast runoff based on the basin characteristics of Fengman reservoir.Initial weight is optimized by genetic algorithm according to the blindness of random operation of network weight.Through the test of historical floods,the practical value and precision of the model are approved in Fengman reservoir.

关 键 词：洪水预报 BP神经网络 遗传算法 

分 类 号：P338[天文地球—水文科学]