若干水文预报方法综述  

Review on some methods for hydrological forecasting

作　　者：王文[1] 马骏[2] 

机构地区：[1]河海大学水资源环境学院,江苏南京210098 [2]黄河水利委员会水文局,河南郑州450004

出　　处：《水利水电科技进展》2005年第1期56-60,共5页Advances in Science and Technology of Water Resources

摘　　要：将现有水文预报方法分为过程驱动模型方法和数据驱动模型方法两大类.过程驱动模型指以水文学概念为基础,对径流的产流过程与河道演进过程进行模拟,从而进行流量过程预报的模型.过程驱动模型近年在中长期预报方面的发展主要表现在对概念性流域降雨径流模型的结构进行改进,以适应较大时间尺度预报的需要.数据驱动模型则是基本不考虑水文过程的物理机制,而以建立输入输出数据之间的最优数学关系为目标的黑箱子方法.数据驱动模型以回归模型最为常用,近年来由于神经网络模型、非线性时间序列分析模型、模糊数学方法和灰色系统模型等的引进,以及水文数据获取能力和计算能力的发展,数据驱动模型在水文预报中受到了广泛的关注.The current methods for hydrological forecasting are divided into two classes, i.e. the process-driven model and the data-driven model. The process-driven model is based on the conception of hydrology, with which the discharge forecasting can be performed by simulation of the runoff variation and river channel evolution. The advances of process-driven models in medium- and long-term forecasting mainly concentrate on the modification of the precipitation and runoff models of river basins, so that the models can meet the requirement of the medium- and long-term forecasting. While, the data-driven model, without requirement of the analysis of the physical mechanics, is fundamentally a black-box model with an objective of identification of the optimal mathematical relationship between inputs and outputs. Among all the data-driven models, the linear regression model is the most commonly used. Owing to the introduction of some new forecasting methods into hydrological forecasting, such as the artificial neural network model, the nonlinear time-series analysis model, the fuzzy mathematic model, the grey system model, and so on, and the improvement of the capability of data acquisition and calculation, the data-driven model has drawn wide attentions in hydrological forecasting.

关 键 词：河流流量 水文预报 水文模型 过程驱动模型 数据驱动模型 

分 类 号：P338[天文地球—水文科学]