乔广超,杨明祥,刘琦,张洋

• 1:南昌航空大学软件学院
• 2:中国水利水电科学研究院流域水循环模拟与调控国家重点实验室
• 3:北京市城市河湖管理处

摘要(Abstract)：

丹江口水库位于汉江上游,是南水北调中线工程水源地,因受华西秋雨影响经常发生较大的秋汛过程。目前应用于丹江口水库径流预报的方法普遍存在预报精度不高和稳定性不强等缺点。针对上述问题,构建了PSO-SVR中长期预报模型,利用粒子群优化算法(PSO)寻找支持向量回归机(SVR)的惩罚系数C、不敏感系数ε以及高斯径向基核函数的gamma参数,在此基础上利用分层感知器人工神经网络(ANN)分析了SVR预报误差规律和特征,建立了PSO-SVR-ANN耦合模型,实现了径流预报的自纠正。结果显示,PSO-SVR-ANN秋汛期预报模型平均相对误差较小,均在10%左右;合格率较高,均处于80%以上水平。实验结果表明,PSO-SVR-ANN模型与PSO-SVR相比,预报精度更高,稳定性更强,可信度更高,具有一定的实用价值,为相关研究提供了参考。

关键词(KeyWords)： 丹江口水库;秋汛期;径流预报;神经网络;误差修正

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金青年基金资助项目(51709271);国家自然科学基金联合基金项目(U1865102)

作者(Author): 乔广超,杨明祥,刘琦,张洋

DOI: 10.13928/j.cnki.wrahe.2021.04.007

参考文献(References)：

• [1] 刘大海.中长期水文预报方法的分析与展望[J].黑龙江科技信息，2016(13):240.LIU Dahai.Analysis and Prospect of medium and long term hydrological forecasting methods.Heilongjiang Science and Technology Information,2016(13):240.
• [2] 范钟秀.中长期水文预报[M].南京：河海大学出版社，1999.FAN Zhongxiu.Medium and long term hydrological forecast[M].Nanjing:Hohai University Press,1999.
• [3] 贾军伟，张利平，刘恋，等.长江干流站中长期径流预报方法研究[J].水资源研究，2014,3(4):7.JIA Junwei,ZHANG Liping,LIU Lian,et al.Research on Medium and Long Term Runoff Forecast in Yangtze River Basin[J].Journal of Water Resources Research,2014,3(4):7.
• [4] 卢迪，周惠成.基于互信息量与BP神经网络的中长期径流预报方法研究[J].水文，2014,34(4):8-14+67.LU Di,ZHOU Huicheng.Study on the medium and long term runoff forecast method based on mutual information and BP neural network[J].Journal of China Hydrology,2014,34(4):8-14+67.
• [5] 崔东文.改进Elman神经网络在径流预测中的应用[J].水利水运工程学报，2013(2):75-81.CUI Dongwen.An improved Elman neural network and its application to runoff forecast[J].Hydro-Science and Engineering,2013(2):75-81.
• [6] MOHAMAD J A,MOHAMAD R K,OZGUR K,et al.A new approach for simulating and forecasting the rainfall-runoff process within the next two months[J].Journal of Hydrology,2017,548:588-597.
• [7] 杨道辉，马光文，刘起方，等.基于粒子群优化算法的BP网络模型在径流预测中的应用[J].水力发电学报，2006(2):67-70.YANG Daohui,MA Guangwen,LIU Qifang,et al.Application of BP network model based on particle swarm optimization algorithm in runoff prediction[J].Journal of Hydroelectric Engineering,2006(2):67-70.
• [8] 崔东文，金波.动态自适应粒子群优化算法与最小二乘支持向量机在年径流预测中的应用[J].人民珠江，2016,37(10):27-33.CUI Dongwen,JIN Bo.Application of dynamic adaptive particle swarm optimization algorithm and least squares support vector machine in annual runoff prediction[J].Pearl River,2016,37(10):27-33.
• [9] 刘勇，王银堂，陈元芳，等.丹江口水库秋汛期长期径流预报[J].水科学进展，2010,21(6):771-778.LIU Yong,WANG Yintang,CHEN Yuanfang,et al.Long term runoff forecast of Danjiangkou Reservoir in autumn flood season[J].Advances in Water Science,2010,21(6):771-778.
• [10] 白雪梅，赵松山.对主成分分析综合评价方法若干问题的探讨[J].统计研究，1995(6):47-51.BAI Xuemei,ZHAO Songshan.Discussion on the comprehensive evaluation method of principal component analysis[J].Statistical Research,1995(6):47-51.
• [11] HARRIS D,CHRIS J C B,LINDA K,et al.Support Vector Regression Machines[J].Advances in Neural Information Processing Systems,1997,28(7):779-784.
• [12] 王健峰，张磊，陈国兴，等.基于改进的网格搜索法的SVM参数优化[J].应用科技，2012,39(3):28-31.WANG Jianfeng,ZHANG Lei,CHEN Guoxing,et al..SVM parameter optimization based on improved grid search method[J].Applied Science and Technology,2012,39(3):28-31.
• [13] 周驰，高海兵，高亮，等.粒子群优化算法[J].计算机应用研究，2003(12):7-11.ZHOU Chi,GAO Haibing,GAO Liang,et al.Particle swarm optimization[J].Application Research of Computers,2003(12):7-11.
• [14] KUO-LIN H,HOSHIN V G,SOROOSH S.Artificial Neural Network Modeling of the Rainfall-Runoff Process[J].Water Resources Research,1995,31(10):2517-2530.
• [15] 周祥全，张津.深层网络中的梯度消失现象[J].科技展望，2017,27(27):284.ZHOU Xiangquan,ZHANG Jin.The phenomenon of gradient vanish in deep network[J].Science And Technology,2017,27(27):284.
• [16] NITISH S,GEOFFREY E H,ALEX K,et al.A simple way to prvent neural networks from overfitting[J].Journal of Machine Learning Research,2014,15(1),1929-1958.
• [17] IMRAN K M J,AMELIA R I.Adam optimization algorithm for wide and deep neural network[J].Knowledge Engineering and Data Science,2019,2(1):41-46.
• [18] 顾思南，杨修群，GUSINAN,等.北半球绕极涡的变异及其与我国气候异常的关系[J].气象科学，2006,26(2):135-142.GU Sinan,YANG Xiuqun,GU Sinan,et al.Group of northern circumpolar vortex variation and its relationship with climate anomalies in China[J].Meteorological Sciences,2006,26 (2):135-142.
• [19] 陈昊荣，金生.基于SWAT模型的汉江流域径流模拟[J].水利与建筑工程学报，2020(1):245-250.CHEN Haorong,JIN Sheng.Runoff simulation of Hanjiang River Basin based on SWAT model[J].Journal of Water Resources and Architectural Engine,2020(1):245-250.
• [20] 水利部水利信息中心.水文情报预报规范(SL250-2000)[S].北京：中国水利水电出版社，2000.Water Resources Information Center.Standard for hydrological information and hydrological forecasting (SL250-2000)[S].Beijing:China Water & Power Press,2000.