摘要
针对郁江流域灾害性洪水发生频繁,选取南宁水文(2)站历史观测中具有代表性的24场超73 m的洪水水位流量资料进行水位(H)—流量(Q)关系曲线拟合研究,分别采用简单趋势线、多项式及BP神经网络三种不同类型方法来进行拟合。结果表明,三种模拟方法均能满足精度要求,其中三阶正交多项式拟合精度最佳,均方残差达到0.000 001,BP神经网络表现最差,但模型本身不存在明显的优劣之分。因此对于不同流域应该选用何种模型模拟,需结合流域自然地理特征及洪水成因规律进行具体分析。
Disaster flood occurred frequently in Yujiang river basin.This article selects 24 representative over 73 m flood water-level data at Nanning hydrology station(2)to fitting the flow-water level curve.Fitting methods include trend line,polynomial and BP artificial neural network.The results show that the simulation precision of three methods can meet the requirements;the third-order orthogonal polynomial fitting precision is the best and the mean squared residuals is 0.000 001;the precision of BP neural network is the worst,but there is no obvious advantages and disadvantages for three models.For choosing suitable model of simulation basin,it needs to consider the natural geographical characteristics of the basin and specific analysis of flood cause rules.
出处
《水电能源科学》
北大核心
2017年第7期89-91,197,共4页
Water Resources and Power
基金
国家自然科学基金项目(51369005)
广西防灾减灾与工程安全重点实验室系统性研究项目(2013ZDX04)
