Application of ANN and ANFIS models for reconstructing

Hydrological Sciences Journal, 2009

Appropriate outflow from a barrage should be maintained to avoid flooding on the downstream side during the rainy season. Due to the nonlinear and fuzzy behaviour of hydrological processes, and in cases of scarcity of relevant data, it is difficult to simulate the desired outflow using physically-based models. Artificial intelligence techniques, namely artificial neural networks (ANN) and an adaptive neurofuzzy inference system (ANFIS), were used in the reported study to estimate the flow at the downstream stretch of a river using flow data for upstream locations. Comparison of the performance of ANN and ANFIS was made by estimating daily outflow from a barrage located in the downstream region of Mahanadi River basin, India, using daily release data from the Hirakud Reservoir, located some distance upstream of the barrage. To obtain the best input-output mapping, five different models with various input combinations were evaluated using both techniques. The significance of the contribution of two upstream tributaries to barrage outflow estimation was also evaluated. Three feed-forward back-propagation training algorithms were used to train the models. Standard performance indices, such as correlation coefficient, index of agreement, root mean square error, modelling efficiency and percentage deviation in peak flow, were used to compare the performance of the models, as well as the training techniques. The results revealed that the neural network with conjugate gradient algorithm performs better than Levenberg-Marquardt and gradient descent algorithms. The model which considers as input the reservoir release up to three antecedent time steps produced the best results. It was found that barrage outflow could be better estimated by the ANFIS than by the ANN technique. Key words neural networks; fuzzy inference system; hydrological processes; training algorithms Réseau de neurones et systèmes d'inférence neuro-floue adaptatif pour la prévision de débit en rivière Résumé Le flux sortant d'un barrage devrait être maintenu à un niveau approprié pour éviter les inondations à l'aval durant la saison humide. En raison du comportement non-linéaire et flou des processus hydrologiques, et dans le cas où les données pertinentes sont rares, il est difficile de simuler le flux sortant souhaitable à l'aide de modèles à bases physiques. Des techniques d'intelligence artificielle, en l'occurrence à base de réseau de neurones artificiels (RNA) et de système d'inférence neuro-floue adaptatif (ANFIS), ont été utilisées dans cette étude pour estimer le débit dans le tronçon aval d'une rivière à l'aide des données de débit en des sites amont. La comparaison des performances des RNA et de ANFIS a été menée pour l'estimation du flux journalier issu d'un barrage situé dans la région aval du basin de la Rivière Mahanadi en Inde, à partir des données de lâchure du Barrage Hirakud situé à l'amont. Afin d'obtenir la meilleure cartographie entrées-sorties, cinq modèles différents avec plusieurs combinaisons d'entrées ont été évalués avec les deux techniques. La significativité de la contribution des deux affluents amont pour l'estimation du flux sortant du barrage a également été évaluée. Trois algorithmes d'apprentissage progressif avec rétropropagation ont été utilisés pour renseigner les modèles. Des indices de performance standard, comme le coefficient de corrélation, l'indice de satisfaction, l'erreur quadratique moyenne, l'efficience de modélisation et le pourcentage d'écart du pic de débit, ont été utilisés pour comparer les performances des modèles ainsi que des techniques d'apprentissage. Les résultats montrent que le réseau de neurones avec un algorithme de gradient conjugué donne de meilleurs résultats que les algorithmes de Levenberg-Marquardt et de gradient descendant. Le modèle qui considère comme entrées les lâchures lors des trois pas de temps antérieurs donne les meilleurs résultats. Le flux sortant du barrage est mieux estimé avec la technique ANFIS qu'avec les RNA. Mots clefs réseaux de neurones; système d'inférence floue; processus hydrologiques; algorithms d'apprentissage Open for discussion until

2014

Investigation of the level of water supplies is considered as a very signifi cant and prominent procedure in basin area programs. Studying the groundwater resources is much more important in regions suffering from the lack of surface waters.In this study, three methods of artificial neural networks, Adaptive Neuro - Fuzzy Inference S ystem and Time Series havebeen assessed in order to find the best way to predict ground water levels in North Mahyar plain , Isfahan. The rainfall, temperature, relative humidity, the operation wells and aquifer fed by the near aquifer are considered as inp ut data, and groundwater levels of 14 observed wells are considered as output. The results showed that the Adaptive Neuro Fuzzy Inference System can give more accuracy for predicting groundwater level than Time Series analysis and artificial

The frequency of occurrence of extreme hydrologic event like flood is important in water resources planning and management. Though there is a definite relationship between the frequency of occurrences and magnitude of the extreme event, reliable prediction of maximum discharge remains as a challenge due to uncertainties and non-linearity of influencing parameters. Statistical techniques are commonly used for finding the maximum discharge and return period relationship. However, these techniques are generally considered to be inadequate because of the non-linearity of the problem. In this study, artificial neural network and adaptive neuro-fuzzy inference system are employed in order to capture the non-linear relationship between annual maximum discharge and frequency. The developed models are validated using Godavari River basin data. Performances of the developed models were compared with respect to root mean square errors, efficiency and coefficient of determination. Based on these results, it was found that artificial neural network performs marginally better than that of adaptive neuro-fuzzy inference system.

vzb.baw.de

Although suspended sediment load can be predicted using numerous developed equations their results often differ from each other and from measured data due to complexity of sediment transport nature. In recent years, simulation models for prediction of suspended sediment load of rivers have been popular among researchers because of Progress of computer models. Artificial Neural Networks (ANN) and Artificial Neuro-Fuzzy Inference Systems (ANFIS) are two well-known models for prediction of hydraulic and hydrology events. Many researchers have studied the application of Artificial Neural Networks in vital topics of hydrology and hydraulics such as prediction of sediment load, rainfall-runoff modeling, flow prediction etc. Cigizoglu (2002) made a comparison between ANNs and SRC for suspended sediment estimation and found that the estimations obtained by ANN's were significantly superior to the corresponding classical sediment rating curve ones. Agarwalet et al (2006) simulated the runoff and sediment yield using artificial neural networks as daily, weekly, ten-daily, and monthly monsoon runoff and sediment yield from an Indian catchment using back propagation artificial neural network (BPANN) technique, and compared the results with observed values obtained from using single-and multi-input linear transfer function models. They showed that the ANN model gives pretty reliable results. Kisi (2005) investigated the abilities of neuro-fuzzy (NF) and neural network (NN) approaches to model the stream flowsuspended sediment relationship for two stations-Quebrada Blanca station and Rio Valenciano station-operated by the US Geological Survey. He found that the NF model gives better estimates than the other technique. The scope of this study is the suspended sediment estimation of Ekbatan dam using an intelligent method to get more accurate results compared to the rating curve. Two ANN and ANFIS algorithms are trained using measured water and sediment discharge data of Yalfan gauging station which is located at the entrance of Ekbatan dam in Iran. Dependency of suspended

International Journal of Engineering Research and, 2021

Missing data has been a common problem and has been confronted by many researchers in the field of hydrology. Rainfall and Temperature time series data are often found missing and such missingness have huge implication on hydrological modelling, flood frequency analysis, trend analysis and dam operation schemes. Owing to the presence of missing data it hinders the performance analysis of the data and inhibits in concluding the correct inferences from the data. In this study, missing data in the rainfall and temperature has been imputed using kNN model and Tree-based model and subsequently these imputed data have been used as predictors to predict the river flow data using Artificial Neural Network (ANN). Uncertainty from kNN imputation model has been found with bootstrapping techniques, while the tree based and ANN model were assessed by Root Mean Square Error (RMSE) and Mean Absolute Error

The Scientific World Journal, 2015

Water table forecasting plays an important role in the management of groundwater resources in agricultural regions where there are drainage systems in river valleys. The results presented in this paper pertain to an area along the left bank of the Danube River, in the Province of Vojvodina, which is the northern part of Serbia. Two soft computing techniques were used in this research: an adaptive neurofuzzy inference system (ANFIS) and an artificial neural network (ANN) model for one-month water table forecasts at several wells located at different distances from the river. The results suggest that both these techniques represent useful tools for modeling hydrological processes in agriculture, with similar computing and memory capabilities, such that they constitute an exceptionally good numerical framework for generating high-quality models.

JNSPS, 2023

The study proposed an Adaptive Neuro-Fuzzy Inference Systems (ANFIS) model capable of predicting sediment deposited in a dam and sediment loss-in-transit (SLIT) using the potential of a formulated mathematical relation. The input parameters consists of five members viz: the rainfall, the slope, the particle size, the velocity, and the computed total volume of sediment exited from two prominent gullies for 2017, 2018, and 2019. The outputs are the total volume of sediment deposited at the adjoining Ikpoba dam for 2017, 2018, and 2019, respectively. The Ordinary Least Square (OLS) regression model on sediment volume retained all covariates with p<0.05, explaining 93.8% of the variability in the dataset. The multicollinearity effect on the dataset was assessed using the Variance Inflation Factor (VIF) which was found not to pose a problem for (VIF<5). The model was validated using the (MSE), the (MAE), and the correlation coefficient (r). The best prediction was obtained as: (RMSE = 0.0423; R 2 = 0.947). The predicted volume of sediment was 842,895.8547m 3 with an error of-0.3295344% and the predicted volume of SLIT was 57,787.98m 3 which is an indication that ANFIS performs satisfactorily in predicting sediment volume for the gullies and the dam respectively.

2011

R iver level forecasting is quite important for reservoir operation studies, flood planning and control, modeling and management water resources. In the last decade, the softcomputing model as a branch of the artificial intelligence science were introduction as a forecast tool beside knowledge based system, expert system, fuzzy logic, artificial neural network, and genetic algorithm. The method that used in this research was a combination between fuzzy logic and artificial neural network which usually called neuro fuzzy system of adaptive neuro fuzzy inference system (ANFIS) algorithm approach was used construct a river level forecasting system. The advantages of this method is that is use input- output data sets. In particular, the applicability of ANFIS as an estimation model for river flow was investigated. To illustrate the applicability and capability the ANFIS, the River Indragir i , located the Indragiri Hulu Residence and the most important water resources of Indragiri c ...

2017

Disasters caused by floods are a major cause of losses of properties and lives. The unpredictability in weather conditions due to changing weather patterns do not only lead to flooding but also contribute to water resource management problems. Rapid development in many tropical countries, like in Malaysia, has resulted in the loss of natural floodplains leading to an increase in flooding and water shortage. Sufficient advanced flood warning system that can save lives and properties can be developed using accurate river model. The work reported in this thesis has made significant vi 3.5 Neural network paradigms. ………………………………………………. 3.6 Evaluation criteria. …………………………………………………………….. 3.7 Neural network training. ………………………………………………………. 3.8 ANN application in prediction model. ……………………………………… 3.9 Implementation………………………………………………………………… 3.10 Model Evaluation……………………………………………………………… 3.11 Correlation Coefficient……………………………………………………….. 3.12 Advantages and Disadvantages of ANN for Hydrological Modelling….. 3.13 Summary………………………………………………………………………… 3.14 Major Outcome……………………………………………………………… CHAPTER 4 METHODOLOGY 4.1 Study area ……………………………………………………………………… 4.2 Mann-Kendall (MK) trend test……………………………………………… 4.3 Rain Dynamics…………………………………………………………………… 4.4 Catchment Area Characteristic………………………………………………… 4.5 Surface-runoff relationship………………………………………………. 4.6 ANN-based river flow rate model ……………………………………………. 4.7 Comparison between river level and river flow rate measurement ……….. 4.8 Summary………………………………………………………………………… 4.9 Major Outcome …………………………………………………………………. CHAPTER 5. RESULTS 5.1 Rainfall statistics. …………………………………………………………… 5.2 Comparison between level monitoring and flow monitoring for flood observation……………………………………………………………………… 5.2.1 Case 1: River Pelarit Flood event on 30th October to 1st November 2010. 5.2.2 Case 2: River Jarum Flood event on 30th October to 1st November 2010…. 5.2.3 Case 3: River Pelarit, Flood event on 1 April 2011…………………………. 5.2.4 Case 4: River Jarum, Flood event on 1 April 2011……………………………. vii 5.2.5 Summary for comparison between level monitoring and flow measurement 5.3 Rainfall in the catchment area and river flow rate ………..…………………. 5.3.1 Dry period before rain against time delay to change flow rate…………… 5.3.2 Rain-No change condition……………………………………………………… 5.4 ANN-based River Flow Model………………………………………………… 5.4.1 Model I: Flow Rate Model for River Pelarit using LM Training Algorithm. 5.4.2 Model II River Flow Rate Model for Pelarit using BR Algorithm …………. 5.4.3 Model III River Flow Rate for River Pelarit using LM Training Algorithm. 5.4.4 Model IV River Flow for River Pelarit using BR Training Algorithm……..

Advances in Water Resources, 2006

Accurate prediction of the water level in a reservoir is crucial to optimizing the management of water resources. A neuro-fuzzy hybrid approach was used to construct a water level forecasting system during flood periods. In particular, we used the adaptive network-based fuzzy inference system (ANFIS) to build a prediction model for reservoir management. To illustrate the applicability and capability of the ANFIS, the Shihmen reservoir, Taiwan, was used as a case study. A large number (132) of typhoon and heavy rainfall events with 8640 hourly data sets collected in past 31 years were used. To investigate whether this neuro-fuzzy model can be cleverer (accurate) if human knowledge, i.e. current reservoir operation outflow, is provided, we developed two ANFIS models: one with human decision as input, another without. The results demonstrate that the ANFIS can be applied successfully and provide high accuracy and reliability for reservoir water level forecasting in the next three hours. Furthermore, the model with human decision as input variable has consistently superior performance with regard to all used indexes than the model without this input.