混沌序列WA-VOLTERRA耦合模型在月降水量预测中的应用

摘要: 由于月降水量时间序列含有大量噪声, 并表现出明显的混沌特性, 现有预测模型均存在一定程度的不足.基于混沌理论的小波分析-VOLTERRA级数自适应(WA-VOLTERRA)耦合预测模型, 在对月降水量时间序列进行混沌特性识别的基础上, 先用小波分析对月降水序列进行时频分解, 再分别对各频率分量进行相空间重构并用3阶VOLTERRA级数自适应模型建模预测, 最后综合得到原始序列的预测值.以相近区域杭州市和南通市的月降水序列为例, 并通过与小波分析-支持向量机(WA-SVM)模型进行比较, 发现该模型具有较强的适用性和更高的预测精度.
- 月降水时间序列 /
- 混沌理论 /
- 小波分解 /
- VOLTERRA级数自适应模型
Abstract: To address the inefficiency of exsiting prediction models of monthly precipitation time series due to large amount of noises and obvious characteristics of chaos, a coupling model is proposed in this study, which takes full advantages of wavelet analysis and VOLTERRA adaptive model. The monthly precipitation time series is firstly mapped into several time-frequency domains, and then a third-order VOLTERRA adaptive model is established for each domain based on the phase-space reconstruction. The final forecasting results are the algebraic sums of all the forecasted components obtained by respective VOLTERRA adaptive model corresponding to different time-frequency domains. An experiment has been conducted by applying different models to estimate the monthly precipitation time series in Hangzhou and Nantong, and the comparison of the data obtained by the conventional model with the results obtained using wavelet analysis and support vector machine (WA-SVM) coupling prediction model confirms that this new WA-VOLTERRA coupling method can achieve higher accuracy. The new model offers a new approach for monthly precipitation forecasting.
- monthly precipitation time series /
- chaos theory /
- wavelet analysis /
- VOLTERRA adaptive model

HTML全文

图 1 WA-VOLTERRA自适应耦合预测模型示意图

Fig. 1. The diagram of WA-VOLTERRA adaptive coupling forecast model

下载: 全尺寸图片幻灯片

图 2 杭州市月降水序列log₂(r)-log₂c(r)关系

Fig. 2. Relational curves of log₂(r)-log₂c(r) for monthly precipitation series of Hangzhou

下载: 全尺寸图片幻灯片

图 3 杭州市月降水序列饱和关联维图

Fig. 3. Relational curves of saturation correlation dimension for monthly precipitation series of Hangzhou

下载: 全尺寸图片幻灯片

图 4 小数据量法计算得到的Lyapunov值

Fig. 4. The calculation chart of Lyapunov value

下载: 全尺寸图片幻灯片

图 5 小波分解后的各个分量序列

Fig. 5. Wavelet decomposition of the monthly precipitation of Hangzhou City

下载: 全尺寸图片幻灯片

图 6 杭州市月降水量预测值与实测值对比

Fig. 6. One month ahead precipitation forecasts using WA-VOLTERRA model and WA-SVM model of Hangzhou

下载: 全尺寸图片幻灯片

图 7 南通市月降雨量模型预测值与真实值对比

Fig. 7. The comparison of Nantong monthly precipitation predicted and measured values

下载: 全尺寸图片幻灯片

表 1 杭州市月降水量预测结果对比

Table 1. Comparison of different models for 1 month ahead precipitation forecasting at Hangzhou

模型 RMSE MAPE(%)

WA-VOLTERRA 22.9 24.5

WA-SVM 28.2 30.1

下载: 导出CSV

表 2 南通市模型预测结果对比

Table 2. Comparison of different models for 1 month ahead precipitation forecasting at Nantong

模型 RMSE MAPE(%)

WA-VOLTERRA 16.9 31.8

WA-SVM 24.9 36.9

下载: 导出CSV

参考文献(33)

[1]	Chen, C.J., Ni, C.J., 2011. Testing for Nonlinearity in Time Series of Monthly Precipitation in Panxi Region. Plateau and Mountain Meteorology Research, 31(2): 26-30 (in Chinese with English abstract). http://en.cnki.com.cn/Article_en/CJFDTOTAL-SCCX201102004.htm
[2]	Gao, R., Liu, X.H., 2005. Short-Term Load Forecasting Method Based on Support Vector Machine Combined with Wavelet Transform. Journal of Shandong University (Engineering Science), 35(3): 115-118 (in Chinese with English abstract). http://en.cnki.com.cn/Article_en/CJFDTOTAL-WJSY200504017.htm
[3]	Han, M., 2007. Prediction Theory and Method of Chaotic Time Series. China Water & Power Press, Beijing, 28-30 (in Chinese).
[4]	Lai, Y.C., Lerner, D., 1998. Effective Scaling Regime for Computing the Correlation Dimension from Chaotic Time Series. Physics D, 115(1-2): 1-18. doi: 10.1016/S0167-2789(97)00230-3
[5]	Li, H.X., Xu, S.G., Fan, C.R., 2007. Identification of Chaos of Monthly Runoff and Prediction of Runoff Time Series Using Volterra Adaptive Method. Journal of Hydraulic Engineering, 38(6): 760-766 (in Chinese with English abstract).
[6]	Liang, J., Zeng, G.M., Guo, S.L., et al., 2006. Diagnosis of Chaotic Behavior and Forecast Resouces for Monthly Rainfall in Dongting Lake Area. Water Resources and Power, 24(5): 16-19 (in Chinese with English abstract). http://d.wanfangdata.com.cn/Periodical_sdnykx200605005.aspx
[7]	Ma, X.X., Mu, H.Z., Guo, H.F., 2008. Reservoir Monthly Runoff Forecast Model Based on Wavelet-ANFIS Analysis. Water Resources and Power, 26(1): 26-29 (in Chinese with English abstract). http://search.cnki.net/down/default.aspx?filename=SDNY200801008&dbcode=CJFD&year=2008&dflag=pdfdown
[8]	Sivakumer, B., 2004. Chaos Theory in Geophysics: Past, Present and Future. Chaos, Solitons and Fractals, 19(22): 441-462. doi: 10.1016/S0960-0779(03)00055-9
[9]	Sivakumar, B., Berndtsson, R., Olsson, J., et al., 2001. Evidence of Chaos in Rainfall-Runoff Process. Hydrology Science, 46(1): 131-145. doi: 10.1080/02626660109492805
[10]	Sivakumar, B., Jayawardena, A.W., Fernando, T., 2002. River Flow Forecasting: Use of Phase-Space Reconstruction and Artificial Neural Networks Approaches. Journal of Hydrology, 265(1-4): 225-245. doi: 10.1016/S0022-1694(02)00112-9
[11]	Song, X.Y., Zhang, G.D., 2007. Basin Rainfall Series Forecast Based on WA-SVM Combined Model. Journal of Yangtze River Scientific Research Institute, 24(5): 23-26 (in Chinese with English abstract). http://qikan.cqvip.com/Qikan/Article/Detail?id=25715296
[12]	Vallejos, R.O., Anteneodo, C., 2002. Theoretical Estimates for the Largest Lyapunov Exponent of Many-Particle Systems. Physical Review E, 66(2): 1203-1218. doi: 10.1103/PhysRevE.66.021110
[13]	Wang, D.Z., Xia, J., Zhang, L.P., 2002. Chaos Analysis of Monthly Precipitation Time Series in North-East China Area. International Journal Hydroelectric Energy, 20(3): 32-34 (in Chinese with English abstract). http://en.cnki.com.cn/Article_en/CJFDTOTAL-SDNY200203010.htm
[14]	Wang, H.R., Song, Y., Liu, C.M., et al., 2004. Application and Issues of Chaos Theory in Hydroscience. Advances in Water Science, 15(3): 400-407 (in Chinese with English abstract). http://en.cnki.com.cn/Article_en/CJFDTOTAL-SKXJ200403025.htm
[15]	Wei, B.L., Luo, X.S., Wang, B.H., et al., 2002. A Method Based on the Third-Order Volterra Filter for Adaptive Predictions of Chaotic Time Series. Acta Physica Sinica, 51(10): 2205-2210 (in Chinese with English abstract). doi: 10.7498/aps.51.2205
[16]	Yang, Y.G., Chen, Y.H., 2009. Chaotic Characteristics and Prediction for Water Inrush in Mine. Earth Science—Journal of China University of Geosciences, 34(2): 258-262 (in Chinese with English abstract). doi: 10.3799/dqkx.2009.024
[17]	You, R.Y., Chen, Z., Xu, S.C., et al., 2004. Study on Phase-Space Reconstruction of Chaotic Signal Based on Wavelet Transform. Acta Physica Sinica, 53(9): 2882-2888 (in Chinese with English abstract). doi: 10.7498/aps.53.2882
[18]	Yu, G.R., Xia, Z.Q., 2008. Prediction Model of Chaotic Time Series Based on Support Vector Machine and Its Application to Runoff. Advances in Water Science, 19(1): 116-122 (in Chinese with English abstract). http://www.researchgate.net/publication/279617604_Prediction_model_of_chaotic_time_series_based_on_support_vector_machine_and_its_application_to_runoff
[19]	Zhang, J.S., Xiao, X.C., 2000. Predicting Low-Dimensional Chaotic Time Series Using Volterra Adaptive Filers. Acta Physica Sinica, 49(3): 403-408 (in Chinese with English abstract). doi: 10.7498/aps.49.403
[20]	陈超君, 倪长健, 2011. 攀西地区月降水时序非线性特性分析. 高原山地气象研究, 31(2): 26-30. doi: 10.3969/j.issn.1674-2184.2011.02.004
[21]	高荣, 刘晓华, 2005. 基于小波变换的支持向量机短期负荷预测. 山东大学学报(工学版), 35(3): 115-118. doi: 10.3969/j.issn.1672-3961.2005.03.027
[22]	韩敏, 2007. 混沌时间序列预测理论与方法. 北京: 中国水利水电出版社, 28-30.
[23]	李红霞, 许士国, 范垂仁, 2007. 月径流序列的混沌特征识别及Volterra自适应预测法的应用. 水利学报, 38(6): 760-766. doi: 10.3321/j.issn:0559-9350.2007.06.019
[24]	梁婕, 曾光明, 郭生练, 等, 2006. 洞庭湖区月降雨序列的混沌特性识别及预测研究. 水电能源科学, 24(5): 16-19. doi: 10.3969/j.issn.1000-7709.2006.05.005
[25]	马细霞, 穆浩泽, 郭慧芳, 2008. 基于小波-ANFIS的水库月径流预报模型. 水电能源科学, 26(1): 26-29. doi: 10.3969/j.issn.1000-7709.2008.01.007
[26]	宋星原, 张国栋, 2007. 基于WA-SVM组合模型的流域月降雨量预测研究. 长江科学院院报, 24(5): 23-26. doi: 10.3969/j.issn.1001-5485.2007.05.007
[27]	王德智, 夏军, 张利平, 2002. 东北地区月降雨时间序列的混沌特性研究. 水电能源科学, 20(3): 32-34. doi: 10.3969/j.issn.1000-7709.2002.03.011
[28]	王红瑞, 宋宇, 刘昌明, 等, 2004. 混沌理论及在水科学中的应用与存在的问题. 水科学进展, 15(3): 400-407. doi: 10.3321/j.issn:1001-6791.2004.03.025
[29]	韦保林, 罗晓曙, 汪秉宏, 等, 2002. 一种基于三阶Volterra滤波器的混沌时间序列自适应预测方法. 物理学报, 51(10): 2205-2210. https://www.cnki.com.cn/Article/CJFDTOTAL-WLXB200210006.htm
[30]	杨永国, 陈玉华, 2009. 矿井涌水量混沌特征与预测. 地球科学——中国地质大学学报, 34(2): 258-262. https://www.cnki.com.cn/Article/CJFDTOTAL-DQKX200902004.htm
[31]	游荣义, 陈忠, 徐慎初, 等, 2004. 基于小波变换的混沌信号相空间重构研究. 物理学报, 53(9): 2882-2888. doi: 10.3321/j.issn:1000-3290.2004.09.014
[32]	于国荣, 夏自强, 2008. 混沌时间序列支持向量机模型及其在径流预测中应用. 水科学进展, 19(1): 116-122. doi: 10.3321/j.issn:1001-6791.2008.01.020
[33]	张家树, 肖先赐, 2000. 混沌时间序列的Volterra自适应预测. 物理学报, 49(3): 403-408. https://www.cnki.com.cn/Article/CJFDTOTAL-WLXB200003003.htm