Runoff Prediction Based on Deep Belief Network in Decomposition-Correction-Integration Mode

QIAN Yu-xia; CHEN Fu-long; HE Chao-fei; LONG Ai-hua; SUN Huai-wei; L&#xDC; Ting-bo

doi:10.11988/ckyyb.20221671

PDF(11121 KB)

Journal of Changjiang River Scientific Research Institute ›› 2024, Vol. 41 ›› Issue (5) : 35-44. DOI: 10.11988/ckyyb.20221671

Water Resources

Runoff Prediction Based on Deep Belief Network in Decomposition-Correction-Integration Mode

QIAN Yu-xia^1,2, CHEN Fu-long^1,2, HE Chao-fei^1,2, LONG Ai-hua^1,3, SUN Huai-wei^1,4, LÜ Ting-bo^1,2

Author information +

History +

Abstract

Accurate short-term runoff prediction can provide important scientific basis for water resources planning, flood control and drought relief in river basin. To mitigate systematic errors and enhance runoff prediction accuracy of models, we propose the decomposition-correction-integration framework based on the decomposition-integration model. Within this framework, we construct the EEMD-DBN-EnKF and VMD-DBN-EnKF models. Leveraging the Ensemble Kalman Filter data assimilation algorithm, we correct components deviating significantly from measured runoff to alleviate systematic errors introduced by the decomposition process in prediction. Comparative analysis is conducted against the unmodified EEMD-DBN, VMD-DBN, and single DBN models. Results demonstrate that the combination model based on modal decomposition reduces RMSE by a minimum of 23% compared to individual models, while NSE and R² increase by over 21%. Notably, the runoff component-corrected combined model exhibits improved evaluation metrics relative to the modal decomposition-based model. Among these models, the VMD-DBN-EnKF prediction model exhibits the least error and highest effectiveness, with NSE and R² exceeding 0.89, followed by EEMD-DBN-EnKF, VMD-DBN, and EEMD-DBN in descending order. In conclusion, the “decomposition-correction-integration” prediction framework demonstrates robust applicability in the Manas River Basin, offering valuable technical support for short-term runoff forecasts.

Key words

modal decomposition / deep belief network / ensemble Kalman filter / runoff prediction / combination model

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

QIAN Yu-xia, CHEN Fu-long, HE Chao-fei, LONG Ai-hua, SUN Huai-wei, LÜ Ting-bo. Runoff Prediction Based on Deep Belief Network in Decomposition-Correction-Integration Mode[J]. Journal of Changjiang River Scientific Research Institute. 2024, 41(5): 35-44 https://doi.org/10.11988/ckyyb.20221671

References

[1] 周天军,陈梓明,陈晓龙,等.IPCC AR6报告解读:未来的全球气候:基于情景的预估和近期信息[J].气候变化研究进展,2021,17(6):652-663.(ZHOU Tian-jun,CHEN Zi-ming,CHEN Xiao-long,et al. Interpreting IPCC AR6: Future Global Climate Based on Projection under Scenarios and on Near-term Information[J]. Climate Change Research,2021,17(6):652-663.(in Chinese))
[2] 王文, 马骏. 若干水文预报方法综述[J]. 水利水电科技进展, 2005, 25(1): 56-60. (WANG Wen, MA Jun. Review on some Methods for Hydrological Forecasting[J]. Advances in Science and Technology of Water Resources, 2005, 25(1): 56-60.(in Chinese))
[3] 李静, 黄强, 杨元园, 等. 引汉济渭工程调水区月径流预报模型研究[J]. 西安理工大学学报, 2021, 37(3): 338-344. (LI Jing, HUANG Qiang, YANG Yuan-yuan, et al. Research on Monthly Runoff Forecasting Model in Water Diversion Area of the Han to Wei Diversion Project[J]. Journal of Xi’an University of Technology, 2021, 37(3): 338-344.(in Chinese))
[4] SIBTAIN M, LI X, HASSAN B, et al. A Hybrid Model for Runoff Prediction Using Variational Mode Decomposition and Artificial Neural Network[J]. Water Resources, 2021, 48(5): 701-712.
[5] GONG B, SHU C, HAN S, et al. Mine Vegetation Identification via Ecological Monitoring and Deep Belief Network[J]. Plants, 2021, 10(6): 1099.
[6] 胡义明,陈腾,罗序义,等.基于机器学习模型的淮河流域中长期径流预报研究[J].地学前缘,2022,29(3):284-291.(HU Yi-ming, CHEN Teng, LUO Xu-yi, et al. Medium to Long Term Runoff Forecast for the Huai River Basin Based on Machine Learning Algorithm[J]. Earth Science Frontiers, 2022, 29(3): 284-291.(in Chinese))
[7] 张娜. 基于动力系统求非线性优化的局部最优解[J]. 哈尔滨商业大学学报(自然科学版),2015,31(6):744-747.(ZHANG Na.Obtaining Local Optimal Solutions of Nonlinear Programming Problems Based on Gradient System[J]. Journal of Harbin University of Commerce (Natural Sciences Edition),2015,31(6):744-747.(in Chinese))
[8] 李滨, 曾辉. 改进的深度置信网络在电主轴故障诊断中的应用[J]. 机械科学与技术, 2021, 40(7): 1051-1057. (LI Bin, ZENG Hui. Application of Improved Deep Belief Network in Electric Spindle Fault Diagnosis[J]. Mechanical Science and Technology for Aerospace Engineering, 2021, 40(7): 1051-1057.(in Chinese))
[9] WANG L, ZHANG T R, WANG X Y, et al. An approach of Improved Multivariate Timing-Random Deep Belief Net Modelling for Algal Bloom Prediction[J]. Biosystems Engineering, 2019, 177: 130-138.
[10]初海波,魏加华,王东方,等.基于预报因子聚类分级的日径流预报深度信念模型及应用[J].应用基础与工程科学学报,2018,26(5):929-939.(CHU Hai-bo,WEI Jia-hua,WANG Dong-fang,et al. Daily Streamflow Forecasting Using Deep Belief Networks Model Based on Predictors Clustering Classification[J]. Journal of Basic Science and Engineering,2018,26(5):929-939.(in Chinese))
[11]钱立鹏, 刘长征, 陈翠忠, 等. 一种基于深度信念网络的径流量预测方法[J]. 石河子大学学报(自然科学版), 2021, 39(2): 259-264. (QIAN Li-peng, LIU Chang-zheng, CHEN Cui-zhong, et al. A Method of the Runoff Prediction Based on Deep Belief Network[J]. Journal of Shihezi University (Natural Science), 2021, 39(2): 259-264.(in Chinese))
[12]王秀杰,封桂敏,耿庆柱.小波分析组合模型在日径流预测中的应用研究[J].自然资源学报,2014,29(5):885-893.(WANG Xiu-jie,FENG Gui-min,GENG Qing-zhu.Application Research on Combined Models Based on Wavelet Analysis in Prediction of Daily Runoff[J]. Journal of Natural Resources,2014,29(5):885-893.(in Chinese))
[13]SABZEHEE F, NAFISI V, POUR S I, et al. Analysis of the Precipitation Climate Signal Using Empirical Mode Decomposition (EMD) over the Caspian Catchment Area[J]. ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2019, XLII-4/W18: 923-929.
[14]柏玲,陈忠升,王充,等.西北干旱区阿克苏河径流对气候波动的多尺度响应[J].地理科学,2017,37(5):799-806.(BAI Ling, CHEN Zhong-sheng, WANG Chong, et al. Multi-scale Response of Runoff to Climate Change in the Aksu River Basin in Arid Area of the Northwestern China[J]. Scientia Geographica Sinica, 2017, 37(5): 799-806.(in Chinese))
[15]吕晗芳, 赵雪花, 桑宇婷, 等. 基于VMD-LSSVM的月径流预测方法研究[J]. 中国农村水利水电, 2020(8): 166-170, 176. (LÜ Han-fang,ZHAO Xue-hua,SANG Yu-ting,et al. Monthly Runoff Prediction Based on VMD-LSSVM[J].China Rural Water and Hydropower,2020(8):166-170,176.(in Chinese))
[16]XIE T, ZHANG G, HOU J W, et al. Hybrid Forecasting Model for Non-stationary Daily Runoff Series: A Case Study in the Han River Basin, China[J].Journal of Hydrology, 2019, 577: 1-49.
[17]徐赫, 张慧峰, 王磊, 等. 基于经验模式分解卡尔曼滤波的径流预报校正[J]. 水力发电, 2015, 41(1): 18-20. (XU He, ZHANG Hui-feng, WANG Lei, et al. Research on Kalman Filtering Runoff Prediction-correction Method Based on Empirical Mode Decomposition[J]. Water Power, 2015, 41(1): 18-20.(in Chinese))
[18]EVENSEN G. Sequential Data Assimilation with a Nonlinear Quasi-geostrophic Model Using Monte Carlo methods to Forecast Error Statistics[J]. Journal of Geophysical Research: Oceans, 1994, 99(C5): 143-162.
[19]PATIL A, RAMSANKARAN R. Improving Streamflow Simulations and Forecasting Performance of SWAT Model by Assimilating Remotely Sensed Soil Moisture Observations[J]. Journal of Hydrology, 2017, 555: 683-696.
[20]李漫漫, 石朋, 尚艳丽, 等. 基于集合卡尔曼滤波的新安江模型状态变量实时修正方法[J]. 河海大学学报(自然科学版), 2019, 47(3): 209-214. (LI Man-man, SHI Peng, SHANG Yan-li, et al. Real-time Updating Method for the State Variables of Xinanjiang Model Based on Ensemble Kalman Filter[J]. Journal of Hohai University (Natural Sciences), 2019, 47(3): 209-214.(in Chinese))
[21]傅甜甜. EnKF方法在北京市平谷区地下水数值模拟中的应用研究[D]. 北京: 中国地质大学(北京), 2021. (FU Tian-tian. Study on Application of EnKF Method in Groundwater Numerical Simulation in Pinggu District, Beijing[D].Beijing: China University of Geosciences, 2021. (in Chinese))
[22]HUANG S, CHANG J, HUANG Q, et al. Monthly Streamflow Prediction Using Modified EMD-based Support Vector Machine[J]. Journal of Hydrology, 2014, 511: 764-775.
[23]CHU H, WEI J, JIANG Y. Middle- and Long-term Streamflow Forecasting and Uncertainty Analysis Using Lasso-DBN-Bootstrap Model[J]. Water Resources Management, 2021, 35(8): 2617-2632.
[24]SAMANTARAY L, DASH M, PANDA R. A Review on Time-frequency, Time-scale and Scale-frequency Domain Signal Analysis[J]. IETE Journal of Research, 2005, 51(4): 287-293.
[25]刘永伟,王文,刘元波,等.基于EnKF法的径流数据同化对SWAT模型参数优化效果评估[J].河海大学学报(自然科学版),2022,50(2):1-10.(LIU Yong-wei, WANG Wen, LIU Yuan-bo, et al. Evaluation of the Optimization Effect of Streamflow Data Assimilation on SWAT Model Parameters Based on the EnKF Approach[J]. Journal of Hohai University (Natural Sciences), 2022, 50(2): 1-10.(in Chinese))
[26]唐豪, 王晓云, 陈伏龙, 等. 基于ERA5-Land数据集的玛纳斯河径流模拟研究[J]. 地学前缘, 2022, 29(3): 271-283. (TANG Hao, WANG Xiao-yun, CHEN Fu-long, et al. Simulation of Manas River Runoff Based on ERA5-land Dataset[J]. Earth Science Frontiers, 2022, 29(3): 271-283.(in Chinese))
[27]陈伏龙, 杨宽, 蔡文静, 等. 基于GAMLSS模型的水文干旱指数研究: 以玛纳斯河流域为例[J]. 地理研究, 2021, 40(9): 2670-2683. (CHEN Fu-long, YANG Kuan, CAI Wen-jing, et al. Study on Hydrological Drought Index Based on GAMLSS: Taking Manas River Basin as an Example[J]. Geographical Research, 2021, 40(9): 2670-2683.(in Chinese))
[28]解佗. 基于分解集成预测模型框架及多目标对冲机制的水电-光伏联合调度研究[D].西安.西安理工大学,2020.(XIE Tuo. Research on the Integrated Forecasting Model Framework and Multi-objective Hedging Mechanism for Hydropower Photovoltaic Joint Scheduling[D]. Xi’an:Xi’an University of Technology,2020.(in Chinese))
[29]张洪波,余荧皓,孙文博,等. 面向EMD分解的径流分量重构方法对比研究[J].南水北调与水利科技, 2017, 15(1): 60-66,166.(ZHANG Hong-bo, YU Ying-hao, SUN Wen-bo,et al. Comparison among Reconstruction Methods for Empirical Mode Decomposition Components in Runoff Prediction[J].South-to-North Water Transfers and Water Science & Technology, 2017, 15(1): 60-66,166.(in Chinese))
[30]周威帆. 基于f-x域模态分解的探地雷达数据处理与解释[D]. 长春: 吉林大学, 2020. (ZHOU Wei-fan. Data Processing and Interpretation of GPR Based on Mode Decomposition in f-x Domain[D].Changchun: Jilin University, 2020. (in Chinese))