Monthly river flow forecasting using artificial neural network and support vector regression models coupled with wavelet transform

Kalteh, Aman Mohammad

doi:10.1016/j.cageo.2012.11.015

Cited by 171 publications

(72 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The nonlinear nodes of the hidden layer are centered so that each of them is specialized on a particular zone of the input space [26,27]. However, the transformation from the hidden layer to the output layer is linear, and it is denoted as the weighted summation of the outputs of all hidden nodes connected to the output nodes:…”

Section: Radial Basis Function Neural Networkmentioning

confidence: 99%

Improved Medium- and Long-Term Runoff Forecasting Using a Multimodel Approach in the Yellow River Headwaters Region Based on Large-Scale and Local-Scale Climate Information

Chen

Wei

et al. 2017

Water

View full text Add to dashboard Cite

Medium-and long-term runoff forecasting is essential for hydropower generation and water resources coordinated regulation in the Yellow River headwaters region. Climate change has a great impact on runoff within basins, and incorporating different climate information into runoff forecasting can assist in creating longer lead-times in planning periods. In this paper, a multimodel approach was developed to further improve the accuracy and reliability of runoff forecasting fully considering of large-scale and local-scale climatic factors. First, with four large-scale atmospheric oscillations, sea surface temperature, precipitation, and temperature as the predictors, multiple linear regression (MLR), radial basis function neural network (RBFNN), and support vector regression (SVR) models were built. Next, a Bayesian model averaging (BMA)-based multimodel was developed using weighted MLR, RBFNN, and SVR models, and the performance of the BMA-based multimodel was compared to those of the MLR, RBFNN, and SVR models. Finally, the high-runoff performance of these four models was further analyzed to prove the effectiveness of each model. The BMA-based multimodel performed better than those of the other models, as well as high-runoff forecasting. The results also revealed that the performance of the forecasting models with multiple climatic factors were generally superior to that without climatic factors. The BMA-based multimodel with climatic factors not only provides a promising, reliable method for medium-and long-term runoff forecasting, but also facilitates uncertainty estimation under different confidence intervals.

show abstract

Section: Radial Basis Function Neural Networkmentioning

confidence: 99%

Improved Medium- and Long-Term Runoff Forecasting Using a Multimodel Approach in the Yellow River Headwaters Region Based on Large-Scale and Local-Scale Climate Information

Chen

Wei

et al. 2017

Water

View full text Add to dashboard Cite

show abstract

“…It is believed that once the noise component of the hydrologic series is appropriately removed using a suitable data-preprocessing technique, the deterministic component can then be easily modeled ). There have been many attempts to address this issue in hydrological and atmospheric data through different techniques (Lisi et al 1995;Sivapragasam et al 2001;Chau and Wu 2010;Wu et al 2010;Wu and Chau 2011;Kalteh 2013Kalteh , 2015Wang et al 2015;Kalteh 2016). For example, Lisi et al (1995) used SSA to extract significant components from the southern oscillation index and applied ANN for forecasting.…”

Section: Introductionmentioning

confidence: 99%

Enhanced Monthly Precipitation Forecasting Using Artificial Neural Network and Singular Spectrum Analysis Conjunction Models

Kalteh

2017

INAE Lett

Self Cite

View full text Add to dashboard Cite

Precipitation is a complex atmospheric process with both temporal and spatial variations which make it difficult to forecast. Several models have been developed to forecast precipitation. This study investigates the accuracy of artificial neural network (ANN) and singular spectrum analysis (SSA) conjunction model in monthly precipitation forecasting. Here, SSA is used to preprocess raw input signals to provide data of high quality to ANN. The conjunction model is obtained by combining two methods, ANN and SSA, and compared with the single/regular ANN. In SSA, the window length and the optimum number of components must be determined. For this purpose, several tentative models, one for each of the available component numbers, are developed, and the related component number of the model with best performance is considered as the optimum number of components. If the selected optimum number is less or equal to the chosen window length which in turn indicates the adequacy of the chosen window length. Monthly precipitation data from one rain gauge station, Ponel station, in northern Iran, is used in the study. The root mean square error (RMSE), correlation coefficient (R) and coefficient of efficiency (CE) statistics are used as the comparing criteria. The comparison of the results reveals that the conjunction models could increase the forecast accuracy of the ANN model in monthly precipitation forecasting. It is found that the conjunction model with RMSE ¼ 52:257, R ¼ 0:858 and CE ¼ 0:731 in the validation/testing period is superior in forecasting monthly precipitations than the most accurate ANN model with RMSE ¼ 91:096, R ¼ 0:444 and CE ¼ 0:183, respectively.

show abstract

“…Also, its computational time was 1.625 s. Thus, the PIO algorithm is an excellent approach to solving the Muskingum model parameter estimation problem. The artificial neural network is a simulation of human physiological structures and mechanisms which depart from symbolic reasoning or logical thinking [30,31]. It plays a role in mitigating imperfect associative memory, defective feature pattern recognition, automatic rule-learning, and other functions.…”

Section: Single-flood Muskingum Model Parameter Estimationmentioning

confidence: 99%

“In-Process Type” Dynamic Muskingum Model Parameter Estimation Method

Zhang

Xie

Zhang

et al. 2017

Water

View full text Add to dashboard Cite

This paper discusses the Muskingum model as a novel parameter estimation method. Sixty representative floods over the past four decades serve as research objects; a linear Muskingum model and Pigeon-inspired optimization (PIO) algorithm are used to obtain the parameters of each flood. The proposed "in-process type" dynamic parameter estimation (IP-DPE) method is used to establish the characteristic attributes set of 50 floods. The characteristic attributes set refers to a set of parameters that could describe the shape, magnitude, and duration of the flood before flood peak; they are the input, whereas parameters K and x of each flood are the output to establish a Neural Network model. Then we input flood characteristic attributes to obtain flood parameters when estimating flood parameters practically. Ten floods were used to test the parameter estimation and flood routing efficacy. The results show that the IP-DPE method can quickly identify parameters and facilitate accurate river flood forecasting.

show abstract

Monthly river flow forecasting using artificial neural network and support vector regression models coupled with wavelet transform

Cited by 171 publications

References 19 publications

Improved Medium- and Long-Term Runoff Forecasting Using a Multimodel Approach in the Yellow River Headwaters Region Based on Large-Scale and Local-Scale Climate Information

Improved Medium- and Long-Term Runoff Forecasting Using a Multimodel Approach in the Yellow River Headwaters Region Based on Large-Scale and Local-Scale Climate Information

Enhanced Monthly Precipitation Forecasting Using Artificial Neural Network and Singular Spectrum Analysis Conjunction Models

“In-Process Type” Dynamic Muskingum Model Parameter Estimation Method

Contact Info

Product

Resources

About