基于BP神经网络的闽江上游洪水预报

摘要

洪水预报是水文科学中的难题,尤其是多分支河流的洪水演进和预报问题更是水文预报的难点.研究应用VB 6.0编程技术,实现人工神经网络BP算法的程序化,并建立闽江上游洪水过程预报的反向传播神经网络模型.经检验,洪水预测精度较好,结果令人满意,为闽江的洪水预报及调度工作提供新的思路和依据..%Flood forecast is one of different problems in hydrologic research, especially in river with several tributaries. With VB 6.0 software , the BP algorithm ( BP for short) of Artificial Neural Network was programmed and the BP flood forecast - model was established for up-river Minjiang district. Through tested, the forecast accuracy is high comparatively with satisfying result. It is shown that simulated stream-flow compares well with the observed data. This provides a new method for flood forecast and flood regulation for Minjiang watershed.