郦于杰 梁忠民 唐甜甜

摘要：根据汉江流域皇庄站1981-2008年逐月径流量与1980-2007年逐月74项环流指数、北太平洋海温场、500 hPa高度场的相关关系，利用逐步回归挑选预报因子，构建基于遗传算法的支持向量回归机模型（GASVR），并对2009-2013年逐月径流量进行预报；结果表明，径流预报精度较高，汛期平均相对误差在30%以内，非汛期、年总量平均相对误差在20%以内，均优于随机森林和多元线性回归模型。将GASVR模型的预报结果作为概率预报的基础，采用贝叶斯理论中的水文不确定性处理器（HUP）对预报的可靠度进行分析；结果表明，HUP不仅可以提供精度更高的定值预报，还能以置信区间的方式量化预报的可靠度，提供更为丰富的预报信息。

关键词：汉江流域；长期径流预报；支持向量回归机；遗传算法；贝叶斯概率预报

中图分类号：P338文献标志码：A文章编号：16721683（2018）03004506

Longterm runoff forecasting based on SVR model and its uncertainty analysis

LI Yujie，LIANG Zhongmin，TANG Tiantian

（Hohai University， College of Hydrology and Water Resources，Nanjing 210098，China）

Abstract：In accordance with the Huangzhuang Station′s monthly runoff from 1981 to 2008 and the correlativity from 1980 to 2007 among the 74 circulation indexes of each month，the monthly north pacific sea surface temperature field，and the 500 hPa geopotential height，we used the stepwise regression method to select the forecast factors and built a GASVR Model （Genetic Algorithm Support Vector Regression Model） on the basis of GA （Genetic Algorithm），in order to forecast the monthly runoff from 2009 to 2013.The results showed that the accuracy of the runoff forecast was relatively high：the average relative error in flood season was within 25%；the yearly runoff amount was within 20% in nonflood season.It was superior to Random Forest and Multiple Regression Model.With the forecast results of the GASVR Model as the basis of the probability forecast，we used the Hydrologic Uncertainty Processor （HUP） of the Bayesian Theory to analyze the forecast reliability.The outcome indicated that HUP could not only give a constantvalue forecast with relatively high accuracy，but also quantify the forecast reliability in the form of a confidence interval to provide more forecast information.

Key words：Hanjiang River basin；longterm runoff forecast；support vector regression；genetic algorithm；bayesian probability forecast

径流的长期预报是指预见期在15 d以上、一年以内，并提供各月径流量的预报，其对防汛抗旱、水资源调度和高效利用具有重要意义[1]。目前，长期径流预报大致可分为物理成因分析法、数理统计法和智能方法三大类[2]。物理成因分析法通过研究陆地海洋下垫面情况、太阳活动、大气环流等要素，推求降水变化规律，再通过水文模型进行径流预报[3]。由于影响径流的因素复杂，该方法实施难度较大，仍处于摸索发展之中。数理统计法根据预报因子类别的不同可分为两种：一是寻求水文要素自身的演变规律进行预报，如根据径流的周期性、趋势性、随机性等特征，采用Morlet小波、方差分析、ARMA等方法构建基于徑流自相关关系的预报模型[47]；……