基于混合核函数SVM水文时序模型及其应用

唐奇, 王红瑞, 许新宜, 王成

系统工程理论与实践 ›› 2014, Vol. 34 ›› Issue (2) : 521-529.

PDF(1282 KB)
PDF(1282 KB)
系统工程理论与实践 ›› 2014, Vol. 34 ›› Issue (2) : 521-529. DOI: 10.12011/1000-6788(2014)2-521
论文

基于混合核函数SVM水文时序模型及其应用

    唐奇, 王红瑞, 许新宜, 王成
作者信息 +

Hydrological time series model based on SVM with mixed kernel function and its application

    TANG Qi, WANG Hong-rui, XU Xin-yi, WANG Cheng
Author information +
文章历史 +

摘要

核函数的选取与构造是SVM应用的关键所在. 传统SVM在水文时序分析方面的应用多是默认选取单一径向基核函数,而忽略了核函数的选择对模型精度和预测结果的影响. 本文基于Mercer核理论,将多项式核与径向基核进行线性组合,构造出混合 核函数,并植入SVM中,对水文时序建立自回归预测模型. 基于武山站和南河川站的月径流预测结果表明,预测序列的相对误差及 均方误差明显优于任一单一核函数. 这是由于混合核函数能够更好地适应并处理复杂的水文时序变化,因此提高了预测精度. 该研究为利用SVM解决复杂多变的非线性水文时序提供了新的探索模式.

Abstract

Kernel function plays an important role in the application of SVM model. Radial basis function (RBF) is used as the most common one in hydrological series analysis and prediction. However, the influence of the selection of kernel function and its structure has been neglected. Therefore, an improved SVM model is developed and used for hydrological predication in this paper. Based on Mercer theorem, polynomial kernel and radial basis kernel are combined to construct the mixed kernel function. Then, auto regression SVM model of mixed kernel function is set up for hydrological series prediction. The model is applied in Wushan and Nanhechuan hydrological station, and the result of monthly runoff prediction shows that the RE and MSE of the improved model are lower than the single kernel function model. This is mainly because mixed kernel function can meet the needs of complex hydrological time series changes, which makes the forecast precision increased. This may also provide a new exploration method for SVM model on research of the non-linear hydrological series.

关键词

支持向量机 / 混合核函数 / 水文时序 / 月径流量 / 武山站 / 南河川站

Key words

support vector machine / mixed kernel function / hydrological series / monthly runoff / Wushan station / Nanhechuan station

引用本文

导出引用
唐奇 , 王红瑞 , 许新宜 , 王成. 基于混合核函数SVM水文时序模型及其应用. 系统工程理论与实践, 2014, 34(2): 521-529 https://doi.org/10.12011/1000-6788(2014)2-521
TANG Qi , WANG Hong-rui , XU Xin-yi , WANG Cheng. Hydrological time series model based on SVM with mixed kernel function and its application. Systems Engineering - Theory & Practice, 2014, 34(2): 521-529 https://doi.org/10.12011/1000-6788(2014)2-521
中图分类号: TV124   

参考文献

[1] Wang H R, Ye L, Liu C M, et al. Problems in wavelet analysis of hydrologic series and some suggestions on improvement[J]. Progress in Natural Science, 2007, 17(1): 80-86.
[2] 桑燕芳, 王栋. 水文时间序列周期识别的新思路与两种新方法[J]. 水科学进展, 2008, 19(3): 412-417.Sang Yanfang, Wang Dong. New idea and two new methods for identifying periodicities of hydrological time series[J]. Advances in Water Science, 2008, 19(3): 412-417.
[3] 王红瑞, 康健, 林欣, 等. 水文序列ARIMA 模型应用中存在的问题与改进方式[J]. 系统工程理论与实践, 2008, 28(10): 166-176.Wang Hongrui, Kang Jian, Lin Xin, et al. Problems existing in ARIMA model of hydrologic series and some improvement suggestions[J]. Systems Engineering-Theory & Practice, 2008, 28(10): 166-176.
[4] 李向阳, 程春田, 林剑艺. 基于BP神经网络的贝叶斯概率水文预报模型[J]. 水利学报, 2006, 37(3): 354-359.Li Xiangyang, Cheng Chuntian, Lin Jianyi. Bayesian probabilistic forecasting model based on BP ANN[J]. Journal of Hydraulic Engineering, 2006, 37(3): 354-359.
[5] 庞博, 郭生练, 熊立华, 等. 改进的人工神经网络水文预报模型及应用[J]. 武汉大学学报:工学版, 2007, 40(1): 33-41.Pang Bo, Guo Shenglian, Xiong Lihua, et al. A modified artificial neural network model and its application to flood forecasting[J]. Engineering Journal of Wuhan University, 2007, 40(1): 33-41.
[6] Yang X H, Mei Y, She D X, et al. Chaotic Bayesian optimal prediction method and its application in hydrological time series[J]. Computers and Mathematics with Applications, 2011, 61: 1975-1978.
[7] Yoon H, Jun S C, Hyun Y, et al. A comparative study of artificial neural networks and support vector machines for predicting groundwater levels in a coastal aquifer[J]. Journal of Hydrology, 2011(1): 128-138.
[8] Tripathi S, Srinivas V V, Nanjundiah R S. Downscaling of precipitation for climate change scenarios: A support vector machine approach[J]. Journal of Hydrology, 2006, 330(3): 621-640.
[9] 黄明聪, 解建仓, 阮本清, 等. 基于支持向量机的水资源短缺风险评价模型及应用[J]. 水利学报, 2007, 38(3): 255-259. Huang Mingcong, Xie Jiancang, Ruan Benqing, et al. Model for assessing water shortage risk based on support vector machine[J]. Journal of Hydraulic Engineering, 2007, 38(3): 255-259.
[10] Wang W C, Chau K W, Cheng C T, et al. A comparison of performance of several artificial intelligence methods for forecasting monthly discharge time series[J]. Journal of Hydrology, 2009(3): 294-306.
[11] Kisi O, Cimen M. A wavelet-support vector machine conjunction model for monthly streamflow forecasting[J]. Journal of Hydrology, 2011(1): 132-140.
[12] 王红瑞, 刘晓红, 唐奇, 等. 基于小波变换的支持向量机水文过程预测[J]. 清华大学学报:自然科学版, 2010, 50(9): 1378-1382.Wang Hongrui, Liu Xiaohong, Tang Qi, et al. Prediction of hydrological series based on wavelet transform and support vector machine[J]. Journal of Tsinghua University: Sci & Tech, 2010, 50(9): 1378-1382.
[13] Noori R, Karbassi A R, Moghaddamnia A, et al. Assessment of input variables determination on the SVM model performance using PCA, Gamma test, and forward selection techniques for monthly stream flow prediction[J]. Journal of Hydrology, 2011(3): 177-189.
[14] 张楠, 夏自强, 江红. 基于多因子量化指标的支持向量机径流预测[J]. 水利学报, 2010, 40(11): 1318-1324.Zhang Nan, Xia Ziqiang, Jiang Hong. Prediction of runoff based on the multiple quantity index of SVM[J]. Journal of Hydraulic Engineering, 2010, 40(11): 1318-1324.
[15] 于国荣, 夏自强. 支持向量机的混沌序列预测模型及在径流中应用[J]. 水利学报, 2007(增刊): 455-460.Yu Guorong, Xia Ziqiang. The prediction model of chaotic series based on support vector machine and its application to runoff[J]. Journal of Hydraulic Engineering, 2007(S): 455-460.
[16] 于国荣, 夏自强. 混沌时间序列支持向量机模型及其在径流预测中应用[J]. 水科学进展, 2008, 19(1): 116-122.Yu Guorong, Xia Ziqiang. Prediction model of chaotic time series based on support vector machine and its application to runoff[J]. Advances in Water Science, 2008, 19(1): 116-122.
[17] 林剑艺, 程春田. 支持向量机在中长期径流预报中的应用[J]. 水利学报, 2006, 37(6): 681-686.Lin Jianyi, Cheng Chuntian. Application of support vector machine method to long-term runoff forecast[J]. Journal of Hydraulic Engineering, 2006, 37(6): 681-686.
[18] Lin G F, Chen G R, Huang P Y, et al. Support vector machine-based models for hourly reservoir inflow forecasting during typhoon-warning periods[J]. Journal of Hydrology, 2009(1): 17-29.
[19] Chen S T, Yu P S, Tang Y H. Statistical downscaling of daily precipitation using support vector machines and multivariate analysis[J]. Journal of Hydrology, 2010(1): 13-22.
[20] 周秀平, 王文圣, 黄伟军. 支持向量机回归模型在径流预测中的应用[J]. 水电能源科学, 2006, 24(4): 4-6.Zhou Xiuping, Wang Wensheng, Huang Weijun. Application of support vector machine regression in runoff forecas[J]. Water Resources and Power, 2006, 24(4): 4-6.
[21] 廖杰, 王文圣, 李跃清, 等. 支持向量机及其在径流预测中的应用[J]. 四川大学学报:工程科学版, 2006, 38(6): 24-28.Liao Jie, Wang Wensheng, Li Yueqing, et al. Support vector machine method and its application to prediction of runoff[J]. Journal of Sichuan University: Engineering Science Edition, 2006, 38(6): 24-28.
[22] 汪丽娜, 李粤安, 陈晓宏. 基于支持向量机的降雨-径流预测研究[J]. 水文, 2009, 29(1): 13-16.Wang Lina, Li Yuean, Chen Xiaohong. Prediction of rainfall-runoff based on support vector machine method[J]. Journal of China Hydrology, 2009, 29(1): 13-16.
[23] 李彦彬, 黄强, 徐建新, 等. 基于混沌支持向量机的河川径流预测研究[J].水力发电学报, 2008, 27(6): 42-47.Li Yanbin, Huang Qiang, Xu Jianxin, et al. Research on prediction of streamflow based on C-SVM[J]. Journal of Hydroelectric Engineering, 2008, 27(6): 42-47.
[24] Yu X Y, Liong S Y. Forecasting of hydrologic time series with ridge regression in feature space[J]. Journal of Hydrology, 2007(3): 290-302.
[25] 李庆国, 陈守煜. 基于模糊模式识别的支持向量机的回归预测方法[J]. 水科学进展, 2005, 16(5): 741-746.Li Qingguo, Chen Shouyu. A SVM regress forecasting method based on the fuzzy recognition theory[J]. Advances in Water Science, 2005, 16(5): 741-746.
[26] 王红瑞, 冯启磊, 林欣, 等. 基于FRCM和FLAR的水文过程遍历特征分析模型及其应用[J]. 中国科学E辑:技术科学, 2008, 38(11): 1984-1994.Wang Hongrui, Feng Qilei, Lin Xin, et al. Development and application of ergodicity model with FRCM and FLAR for hydrological process[J]. Science China: Technological Sciences, 2009, 52(2): 379-386.
[27] 任海军, 孙瑞志, 刘广利. 基于AR_SVR 模型的时间序列预测算法的研究[J]. 计算机工程与设计, 2010, 31(2): 421-424.Ren Haijun, Sun Ruizhi, Liu Guangli. Research of time-series forecasting algorithm based on AR_SVR model[J]. Computer Engineering and Design, 2010, 31(2): 421-424.
[28] ESmith G, Mjordaan E.Improve SVM regression using mixtures of kernels[C]//Proceedings of IEEE International Joint Conference on Neural Networks, USA, 2002: 2785-2790.
[29] 张拥华. 基于支持向量机的金融时间序列研究[D]. 长沙:湖南大学, 2008.Zhang Yonghua. Financial time series researching based on support vector machine[D]. Changsha: Hunan University, 2008.
[30] 水利电力部黄河水利委员会. 1980年黄河流域水文资料第7册, 泾洛谓区(渭河水系)[M]. 开封: 水利电力部黄河水利委员会, 1982.
[31] Jiang G P, Zheng W X. A simple method of chaos control for a class of chaotic discrete-time systems[J]. Chaos, Solitons & Fractals, 2005, 23(3): 843-849.
[32] Chick S, Shortle J, Gelder P V, et al. A model for the frequency of extreme river levels based on river dynamics[J]. Structural Safety, 1996, 18(4): 261-276.
[33] Cline D B H, Pu H H. Verifying irreducibility and continuity of a nonlinear time series[J]. Statistics & Probability Letters, 1998, 40: 139-148.

基金

国家自然科学基金(51279006);国家科技支撑计划(2013BAB05B04)
PDF(1282 KB)

Accesses

Citation

Detail

段落导航
相关文章

/