Ali Azizpor; Ahmad Rajabi; Fariborz Yosefvand; Saeid Shabanlou
Abstract
In the current study, a new hybrid of the genetic algorithm (GA) and adaptive Neuro-fuzzy inference system (ANFIS) was introduced to model the discharge coefficient (DC) of triangular weirs. The genetic algorithm was implemented for increasing the efficiency of ANFIS by adjusting membership functions ...
Read More
In the current study, a new hybrid of the genetic algorithm (GA) and adaptive Neuro-fuzzy inference system (ANFIS) was introduced to model the discharge coefficient (DC) of triangular weirs. The genetic algorithm was implemented for increasing the efficiency of ANFIS by adjusting membership functions as well as minimizing error values. To evaluate the proficiency of the proposed hybrid method, the Monte Carlo simulations (MCS) and the k-fold validation method (k=5) was applied. The results of developed hybrid model indicate that the weir vortex angle, flow Froude number, the ratio of the weir length to its height, the ratio of the channel width to the weir length and ratio of the flow head to the weir height are the most effective parameters in the DC estimation. The quantitative examination of the proposed hybrid method indicates that the Root Mean Square Error (RMSE) and Mean Absolute Percent Error (MAPE) are as 0.016 and 1.647 (respectively) for the superior model. Besides, the Froude number is found as the most effective variable in DC modeling through the quantitative analysis. A comparison of the developed hybrid ANFIS-GA with the individual ANFIS model in the DC estimation indicates the hybrid model outperformed than the individual one.
Mohammad Ali Izadbakhsh; Reza Hajiabadi
Abstract
In the article, through the adaptive neuro-fuzzy inference system (ANFIS), a sensitivity analysis is conducted on the variables affecting the discharge capacity of the weir. To this end, the variables affecting the discharge capacity of labyrinth weirs are initially identified. Then, using these input ...
Read More
In the article, through the adaptive neuro-fuzzy inference system (ANFIS), a sensitivity analysis is conducted on the variables affecting the discharge capacity of the weir. To this end, the variables affecting the discharge capacity of labyrinth weirs are initially identified. Then, using these input parameters, seven ANFIS models are developed for conducting the sensitivity analysis. After that, the most optimal membership function number for the ANFIS model is chosen. In other words, by conducting the trial and error process, the best number of the membership functions in terms of time and modeling accuracy are selected. Then, the sensitivity analysis is performed for the ANFIS models and the superior ANFIS model is chosen finally. The accuracy of the superior model in both the validation and testing artificial intelligence (AI) methods is in an acceptable range. For example, the scatter index (SI), correlation coefficient (R) and the Nash-Sutcliff efficiency coefficient (NSC) for the model in the testing mode are obtained 0.049, 0.964 and 0.924, respectively. It should be noticed that the outcomes of the sensitivity analysis show that the ratio of the weir head to the weir crest and the Froude number are introduced as the most effective input parameters. Eventually, a computer code is proposed to estimate the discharge capacity of labyrinth weirs by this model.
Fariborz Yosefvand; Saeid Shabanlou
Abstract
In this study, for the first time, groundwater level (GWL) variations of the Sarab-e Qanbar well located in the city of Kermanshah, are simulated over a 13-year period by a hybrid model named WANFIS (wavelet-adaptive neuro fuzzy inference system). In order to develop the hybrid model, the wavelet transform ...
Read More
In this study, for the first time, groundwater level (GWL) variations of the Sarab-e Qanbar well located in the city of Kermanshah, are simulated over a 13-year period by a hybrid model named WANFIS (wavelet-adaptive neuro fuzzy inference system). In order to develop the hybrid model, the wavelet transform and the adaptive neuro fuzzy inference system (ANFIS) model are utilized. Furthermore, the 9 and 4 year data are used for training and testing the artificial intelligence models, respectively. Moreover, the effective lags are detected by the autocorrelation function (ACF) and then eight different models are developed for each of the ANFIS and WANFIS models using them. After that, all mother wavelets are evaluated and Dmey mother wavelet is chosen as the most optimal. For this mother wavelet, the values of scatter index (SI), variance account for (VAF) and Root mean square error (RMSE) are obtained 0.192, 94.951 and 3.117, respectively. Next, the superior model is detected through the analysis of the results obtained by all ANFIS and WANFIS models. The superior model estimates the objective function values with reasonable accuracy. For example, the correlation coefficient (R), Scatter Index (SI) and variance account for (VAF) for this model are obtained 0.974, 0.192 and 94.951, respectively. The modeling results indicate that the wavelet transform noticeably enhances the ANFIS model accuracy. Finally, the lags of the time series data for the Sarab-e Qanbar well including (t-1), (t-2), (t-3) and (t-4) are introduced as the most effective lags.
Rahim Gerami Moghadam; Behrouz Yaghoubi; Mohammad Ali Izadbakhsh; Saeid Shabanlou
Abstract
Generally, Hydraulic jumps usually happen at the downstream of hydraulic structures like ogee spillways. In addition, one of the parameters affecting the proper design of stilling basin is calculation of the hydraulic jump length. In this study, a hybrid method (ANFIS-DE) was proposed for modeling hydraulic ...
Read More
Generally, Hydraulic jumps usually happen at the downstream of hydraulic structures like ogee spillways. In addition, one of the parameters affecting the proper design of stilling basin is calculation of the hydraulic jump length. In this study, a hybrid method (ANFIS-DE) was proposed for modeling hydraulic jumps on sloping rough beds for first time. This approach forecasts values of the jump length by combining the Adaptive Neuro-Fuzzy Inference System (ANFIS) and the Differential Evolution (DE) algorithm. First, the variables affecting the hydraulic jump length including the ratio of bed roughness, the Froude number, the ratio of sequent depths and the bed slope were identified. Then, by combining the input parameters, five different numerical models were introduced. Furthermore, the k-fold cross validation (k=4) was utilized so as to verifying the numerical models. The results of the analysis of different numerical models indicated that the model with four input parameters (superior model) simulated the length of the hydraulic jump with higher accuracy. For the best model, the mean absolute percent error (MAPE), the correlation coefficient (R) and the root mean square error (RMSE) were predicted 4.875, 0.978 and 0.807, respectively. Finally, two parameters including the ratio of sequent depths and the Froude number were identified as the most important parameters in modeling the hydraulic jump length on sloping rough beds.
Sultan Noman Qasem; Isa Ebtehaj; Hossien Riahi Madavar
Abstract
Flow through open channels can contain solids. The deposition of solids occasionally occurs due to insufficient flow velocity to transfer the solid particles, causing many problems with transfer systems. Therefore, a method to determine the limiting velocity (i.e. Fr) is required. In this paper, three ...
Read More
Flow through open channels can contain solids. The deposition of solids occasionally occurs due to insufficient flow velocity to transfer the solid particles, causing many problems with transfer systems. Therefore, a method to determine the limiting velocity (i.e. Fr) is required. In this paper, three alternative, hybrid evolutionary algorithm methods, including differential evolution (DE), genetic algorithm (GA) and particle swarm optimization (PSO) based on the adaptive network-based fuzzy inference system are presented: ANFIS-GA, ANFIS-DE and ANFIS-PSO. In these methods, evolutionary algorithms optimize the membership functions, and ANFIS adjusts the premises and consequent parameters to optimize prediction performance. The performance of the proposed methods is compared with that of the general ANFIS using three different datasets comprising a wide range of data. The results show that the hybrid models (ANFIS-GA, ANFIS-DE and ANFIS-PSO) are more accurate than general ANFIS in training with a hybrid algorithm (hybrid of back propagation and least squares). Among the evolutionary algorithms, ANFIS-PSO performed the best (R2=0.976, RMSE=0.26, MARE=0.057, BIAS=-0.004 and SI=0.059).