基于时变权重组合与贝叶斯修正的中长期径流预报

doi:10.18306/dlkxjz.2020.04.011

Abstract

Abstract:

The mid- and long-term runoff prediction with satisfactory accuracy plays an important role as basic information in water resources planning & management and optimal operation of water conservancy projects. Combination and bias reduction are two common post-processing approaches in runoff forecast. Applying them in turn, considering the complicated non-stationary and nonlinear characteristics of runoff, a new mid- and long-term runoff prediction method by connecting time-varying weight combination and Bayesian correction is proposed. This method was used to study the annual and monthly inflow prediction of the Longjiang Reservoir in Yunnan Province. The results show that time-varying weight combination balances the performance difference of the established random forest (RF) and support vector machine (SVM) models in the modeling period and the test period. As a consequence of Bayesian correction, the prediction metrics are close to or better than the best of the predictions of the two individual stages. The proportion of correctly classified hydrological year type reaches 77.2% by employing the forecasted annual runoff, and the Nash-Sutcliffe efficiency coefficient of predicted monthly runoff series is over 0.90. Overall, the method put forward in this study has achieved positive effects in improving the forecast performance.

Key words: mid- and long-term runoff prediction, time-varying weight combination, Bayesian correction, Longjiang Reservoir in Yunnan Province

LI Lingjie, WANG Yintang, HU Qingfang, LIU Yong, LIU Dingzhong, CUI Tingting. Mid- and long-term runoff prediction based on time-varying weight combination and Bayesian correction[J].PROGRESS IN GEOGRAPHY, 2020, 39(4): 643-650.

Figures/Tables 8

Fig.1

Fig.2

Tab.1

Fig.3

Fig.4

Fig.5

Fig.6

Fig.7

References 20

[1]	林沛榕, 张艳军, 冼翠玲 , 等. 不同时间尺度的中长期水文预报研究[J]. 水文, 2017,37(6):2-8.
	[ Lin Peirong, Zhang Yanjun, Xian Cuiling , et al. Time-scale features of medium- and long-range hydrological forecasting. Journal of China Hydrology, 2017,37(6):2-8. ]
[2]	刘勇 . 基于物理成因的中长期水文预报方法与应用研究[D]. 南京: 河海大学, 2011.
	[ Liu Yong . Methods and application for mid-long term hydrological forecast based on physical cause. Nanjing, China: Hohai University, 2011. ]
[3]	陈璐, 叶磊, 卢韦伟 , 等. 基于Copula熵的神经网络径流预报模型预报因子选择[J]. 水力发电学报, 2014,33(6):25-29.
	[ Chen Lu, Ye Lei, Lu Weiwei , et al. Determination of input variables for artificial neural networks for flood forecasting using Copula entropy method. Journal of Hydroelectric Engineering, 2014,33(6):25-29. ]
[4]	赵铜铁钢, 杨大文, 蔡喜明 , 等. 基于随机森林模型的长江上游枯水期径流预报研究[J]. 水力发电学报, 2012,31(3):18-24, 38.
	[ Zhao Tongtiegang, Yang Dawen, Cai Ximing , et al. Predict seasonal low flows in the upper Yangtze River using random forests model. Journal of Hydroelectric Engineering, 2012,31(3):18-24, 38. ]
[5]	Yang T, Asanjan A A, Welles E , et al. Developing reservoir monthly inflow forecasts using artificial intelligence and climate phenomenon information[J]. Water Resources Research, 2017,53(4):2786-2812. doi: 10.1002/2017WR020482
[6]	Tan Q F, Lei X H, Wang X , et al. An adaptive middle and long-term runoff forecast model using EEMD-ANN hybrid approach[J]. Journal of Hydrology, 2018,567:767-780. doi: 10.1016/j.jhydrol.2018.01.015
[7]	Le X H, Ho H V, Lee G , et al. Application of Long Short-Term Memory (LSTM) neural network for flood forecasting[J]. Water, 2019,11(7):1387. doi: 10.3390/w11071387. doi: 10.3390/w11071387
[8]	Chowdhury S, Sharma A . Multisite seasonal forecast of arid river flows using a dynamic model combination approach[J]. Water Resources Research, 2009,45(10):2381-2386.
[9]	Zhang L, Yang X L . Applying a multi-model ensemble method for long-term runoff prediction under climate change scenarios for the Yellow River Basin, China[J]. Water, 2018,10(3):301. doi: 10.3390/w10030301. doi: 10.3390/w10030301
[10]	Men B H, Long R S, Zhang J H . Combined forecasting of streamflow based on cross entropy[J]. Entropy, 2016,18(9):336. doi: 10.3390/e18090336. doi: 10.3390/e18090336
[11]	孙惠子 . 中长期径流的多种组合预测方法及其比较[D]. 杨凌: 西北农林科技大学, 2012.
	[ Sun Huizi . A variety of combination forecasting method and comparing of medium and long-term stream-flow. Yangling, China: Northwest A & F University, 2012. ]
[12]	Wu J H, Qian H, Li P , et al. A system-theory-based model for monthly river runoff forecasting: Model calibration and optimization[J]. Journal of Hydrology and Hydromechanics, 2014,62(1):82-88.
[13]	Liang Z M, Li Y J, Hu Y M , et al. A data-driven SVR model for long-term runoff prediction and uncertainty analysis based on the Bayesian framework[J]. Theoretical and Applied Climatology, 2018,133(1-2):137-149.
[14]	鲁帆, 蒋云钟, 殷峻暹 . 基于BP 神经网络的贝叶斯概率枯季径流预报研究[J]. 水电能源科学, 2008,26(3):7-10.
	[ Lu Fan, Jiang Yunzhong, Yin Junxian . Bayesian probabilistic low flow forecasting research based on BP ANN. Water Resources and Power, 2008,26(3):7-10. ]
[15]	冯小冲 . 水库中长期水文预报模型研究[D]. 南京: 南京水利科学研究院, 2010.
	[ Feng Xiaochong . Study on mid-long term hydrological forecasting model for reservoir. Nanjing, China: Nanjing Hydraulic Research Institute, 2010. ]
[16]	陈华友 . 组合预测方法有效性理论及其应用 [M]. 北京: 科学出版社, 2008: 40-41.
	[ Chen Huayou. The validity theory of combined forecasting method and its application. Beijing, China: Science Press, 2008: 40-41. ]
[17]	Krzysztofowicz R . Bayesian models of forecasted time series 1[J]. Journal of the American Water Resources Association, 1985,21(5):805-814.
[18]	张洪刚 . 贝叶斯概率水文预报系统及其应用研究[D]. 武汉: 武汉大学, 2005.
	[ Zhang Honggang . Probabilistic hydrologic forecasting system based on the Bayesian method. Wuhan, China: Wuhan University, 2005. ]
[19]	李运刚, 何娇楠, 李雪 . 基于SPEI 和SDI 指数的云南红河流域气象水文干旱演变分析[J]. 地理科学进展, 2016,35(6):758-767.
	[ Li Yungang, He Jiaonan, Li Xue . Hydrological and meteorological droughts in the Red River Basin of Yunnan Province based on SPEI and SDI indices. Progress in Geography, 2016,35(6):758-767. ]
[20]	周婷, 金菊良, 李荣波 , 等. 基于小波支持向量机的径流预测性能优化分析[J]. 水力发电学报, 2017,36(10):45-55.
	[ Zhou Ting, Jin Juliang, Li Rongbo , et al. Performance optimization analysis for inflow prediction using wavelet Support Vector Machine. Journal of Hydroelectric Engineering, 2017,36(10):45-55. ]

时段	建模期				检验期
	RF模型		SVM模型		RF模型		SVM模型
	QR	MAPE	QR	MAPE	QR	MAPE	QR	MAPE
年	100	3.83	97.6	6.77	93.3	11.45	93.3	9.39
6月	71.4	15.67	69.0	19.88	66.7	21.29	53.3	20.32
7月	95.2	8.25	83.3	10.81	60.0	22.22	60.0	18.78
8月	100.0	5.56	88.1	10.44	73.3	16.62	93.3	12.51
9月	95.2	7.58	83.3	10.23	80.0	13.02	73.3	16.63
10月	90.5	10.56	81.0	16.87	53.3	24.21	53.3	23.70
11月	90.5	8.75	69.0	11.59	53.3	32.99	40.0	28.66
12月	90.5	8.80	85.7	11.49	60.0	25.13	53.3	22.36
翌年1月	97.6	7.84	85.7	10.80	66.7	20.97	73.3	19.70
翌年2月	95.2	7.14	78.6	9.28	73.3	16.13	80.0	12.11
翌年3月	88.1	9.09	71.4	13.95	60.0	21.51	53.3	25.55
翌年4月	81.0	12.14	57.1	16.93	66.7	18.36	46.7	22.08
翌年5月	71.4	19.24	64.3	24.69	40.0	37.31	66.7	37.99

Mid- and long-term runoff prediction based on time-varying weight combination and Bayesian correction

RichHTML

PDF (PC)

Like

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 8

References 20

Related Articles 0

Metrics

Comments

Recommended 0