WoS İndeksli Yayınlar Koleksiyonu / WoS Indexed Publications Collection
Permanent URI for this collectionhttps://hdl.handle.net/11147/7150
Browse
11 results
Search Results
Now showing 1 - 10 of 11
Article Citation - WoS: 3Citation - Scopus: 3Ensemble and Optimized Hybrid Algorithms Through Runge Kutta Optimizer for Sewer Sediment Transport Modeling Using a Data Pre-Processing Approach(Elsevier, 2023) Safari, Mir Jafar Sadegh; Gül, Enes; Dursun, Ömer Faruk; Tayfur, GökmenUncontrolled sediment deposition in drainage and sewer systems raises unexpected maintenance expenditures. To this end, implementation of an accurate model relying on effective parameters involved is a reliable benchmark. In this study, three machine learning techniques, namely extreme learning machine (ELM), multilayer perceptron neural network (MLPNN), and M5P model tree (M5PMT); and three optimization approaches of Runge Kutta (RUN), genetic algorithm (GA), and particle swarm optimization (PSO) are applied for modeling. The optimization and ensemble hybridization approaches are applied in the modeling procedure. For the case of hybrid optimized models, the ELM and MLPNN models are hybridized with RUN, GA, and PSO algorithms to develop six hybrid models of ELM-RUN, ELM-GA, ELM-PSO, MLPNN-RUN, MLPNN-GA, and MLPNN-PSO. Ensemble hybrid models are developed through coupling the ELM and MLPNN models with the M5PMT algorithm. The data pre-processing approach is applied to find the best randomness characteristic of the utilized data. Results illustrate that the RUN-based hybrid models outperform the GA- and PSO-based counterparts. Although the MLPNN-RUN and MLPNN-M5PMT hybrid models generate better results than their alternatives, MLPNN-M5PMT slightly outperforms MLPNN-RUN model with a coefficient of determination of 0.84 and a root mean square error of 0.88. The current study shows the superiority of the ensemble-based approach to the optimization techniques. Further investigation is needed by considering alternative optimization techniques to enhance sediment transport modeling. © 2023 International Research and Training Centre on Erosion and Sedimentation/the World Association for Sedimentation and Erosion ResearchArticle Citation - WoS: 4Citation - Scopus: 53d Modelling of Surface Spreading and Underground Dam Groundwater Recharge: Egri Creek Subbasin, Turkey(Springer, 2023) Şahin, Yavuz; Tayfur, GökmenThis study investigated surface spreading and underground dam recharge methods to replenish groundwater in Turkey's Egri Creek Sub-basin of the Kucuk Menderes River Basin. A three-dimensional numerical model was employed for this purpose. Field and lab data are provided to the model for realistic simulations. Pumping test results were used to determine the aquifer parameters. The laboratory works involved sieve analysis, permeability tests, and porosity and water content prediction. The numerical model's boundary conditions were determined from the geological and hydrogeological characteristics of the study area. Initial conditions were expressed regarding water content and pressure head in the vadose zone. The numerical model was satisfactorily validated by simulating water levels in three different pumping wells in the study area. Seven different scenarios, each having a different pool size, were investigated for the surface spreading recharge method. The results showed that a pool size of 30 x 30 m with a 6-m depth basin was the most optimal choice, raising the groundwater level to about 29.3 m. On the other hand, it was found that an underground dam could raise the levels by an average of 9.5 m, which might not be significant to warrant the construction.Article Citation - WoS: 13Citation - Scopus: 12Identification of Groundwater Potential Zones in Kabul River Basin, Afghanistan(Elsevier, 2021) Tani, Hamidullah; Tayfur, GökmenGroundwater (GW) plays a vital role in the socio-economic growth of Kabul River Basin (KRB) in Afghanistan. Since the GW resources in the basin have not been properly managed, there is a need for sound strategies by first identifying the potential GW zones. This study assesses the potential groundwater zones for the KRB using the Geographic Information Systems (GIS) and the Analytic Hierarchy Process (AHP). In this direction, seven different thematic maps of rainfall, lithology, land use/land cover, slope, soil, drainage density, and lineament density are first prepared using the GIS. The AHP is then employed to assess the weights of different themes. Finally, the weighted overlay option in the GIS is used to generate the map of the groundwater potential zones (GWPZ). The Very Good zones are mostly located in the downstream and central parts of the KRB, covering around 1543 km(2) area. The Good and the Poor zones are found to be randomly distributed, covering about 39 444 km(2) and 27 658 km(2), respectively. The Very Poor zones are located in the west, southwest, and in some central parts of the basin, covering about 2272 km(2). It is found that only 18% of the total average annual precipitated water of 6.88 x 10(9) m(3)/year infiltrates into the subsurface and ultimately contributes to recharging of the groundwater.Article Citation - WoS: 29Citation - Scopus: 31Reverse Flood Routing in Rivers Using Linear and Nonlinear Muskingum Models(American Society of Civil Engineers, 2021) Badfar, Meisam; Barati, Reza; Doğan, Emrah; Tayfur, GökmenOne of the key factors for flood modeling and control is the flood hydrograph, which is not always available due to lack of flood discharge observations. In reverse flow routing, hydraulic or hydrological calculations are performed from the downstream end to the upstream end. In the present study, a reverse flood routing approach is developed based on the Muskingum model. The storage function is conceptualized as linear and five different nonlinear forms. The Euler and the fourth-order Runge-Kutta numerical methods are used for solving the storage models. The shuffled complex evolution (SCE) algorithm is used for optimization of the flood routing parameters. The models are calibrated and validated with theoretical and actual hydrographs. The results indicate that the proposed methodology could substantially (up to almost 82%) improve comparison with observed inflows. The practical applicability of the proposed methodology is also validated in real river systems.Article Citation - WoS: 1Artificial Neutral Networks To Predict Design Properties for Cemented Embankment Layers of High Speed Train Rail Ways(Foundation Cement, Lime, Concrete, 2013) Egeli, İsfendiyar; Tayfur, Gökmen; Yılmaz, E.; Uşun, HandanI. EGELI, G. TAYFUR, E. YILMAZ, H. USUN ARTIFICIAL NEURAL NETWORKS TO PREDICT DESIGN PROPERTIES FOR CEMENTED EMBANKMENT LAYERS OF HIGH SPEED TRAIN RAILWAYS Cement-Wapno-Beton, Vol. XVIII/LXXX, 2013, No 1, p. 10 High-speed train railway (HSTR) embankment is a complicated process, as it deals with high geometric design standards and material properties. In this study the replaceability of fill strata without cement prepared subgrade layer and with cement addition one is investigated. In the experiments the specimens composed of natural sand with different cement additions and two w/c ratios were used. The Plaxis-FEM (2D) program was employed to find the maximum expected total settlements of HSTR embankments with cemented subgrade layer. Furthermore, the artificial neural networks model was constructed to predict the failure stress, elasticity modulus and strains. The sensivity analysis has revealed that cement content was the most sensitive for stress and elasticity modulus predictions, while the curing age of specimens was for the strain forecast.Article Citation - WoS: 13Citation - Scopus: 14Soil Erosion Model Tested on Experimental Data of a Laboratory Flume With a Pre-Existing Rill(Elsevier Ltd., 2020) Aksoy, Hafzullah; Gedikli, Abdullah; Yılmaz, Murat; Eriş, Ebru; Ünal, N. Erdem; Yoon, Jaeyoung; Tayfur, GökmenPrediction of sediment discharge transported within flow is strongly needed in order to provide measures for a well-established erosion control and water quality management practice. Initiated by runoff generation and erosion processes sediment transport is influenced by microtopography over hillslopes of hydrological watersheds. Consideration of microtopography provides more accurate results. In this study, a process-based two-dimensional rainfall-runoff mathematical model is coupled with erosion and sediment transport component. Both the rainfall-runoff and sediment transport components make simulations in rills and over interrill areas of a bare hillslope. Models at such fine resolution are rarely verified due to the complexity of rills and interrill areas. The model was applied on a data set compiled from laboratory experiments. Erosion flume was filled with granular sand to replace a bare soil. A longitudinal rill and an interrill area were pre-formed over the soil in the flume before the simulated rainfall exerted on. The flume was given both longitudinal and lateral slopes. The simulated rainfall was changed between 45 mm/h and 105 mm/h and exerted on granular uniform fine and medium sand in the erosion flume with longitudinal and lateral slopes both changing from 5% to 20%. Calibration of the model shows that it is able to produce good results in terms of sedigraphs, which suggest also that the model might be considered an important step to verify and improve watershed scale erosion and sediment transport models.Conference Object Upscaling Surface Flow Equations Depending Upon Data Availability at Different Scales(Springer Verlag, 2003) Tayfur, GökmenSt. Venant equations, which are used to model sheet flows, are point-scale, depth-averaged equations, requiring data on model parameters at a very fine scale. When data are available at the scale of a hillslope transect, the point equations need to be upscaled to conserve the mass and momentum at that scale, Hillslope-scale upscaled model must be developed if data are available at that scale. The performance of the three models applied to simulate flows from non-rilled surfaces revealed that the hillslope-scale upscaled model performs as good as the point-scale model though it uses far less data. The transectionally-upscaled model slightly underestimates the observed data.Conference Object Reverse Flood Routing in Rivers(International Association for Hydro-Environment Engineering Research, 2015) Tayfur, Gökmen; Moramarco, TommasoThis study developed model to do riverse flood routing in natural channels. The developed model has basically four components: (1) it expresses an inflow hydrograph by a Pearson Type-III distribution, involving parameters of peak discharge, time to peak, and a shape factor; (2) it employs the basic continutiy equation for flow routing, (3) it relates the storage to downstream flow stage and channel characterictis; and (4) it relates the lateral flow to downstream flow discharge with coefficients. The parameters, coefficients and exponents of the models were obtained using the genetic algorithm method. The developed models are applied to generate upstream hydrographs, using just downstream station information for Ponte Nuovo and Monte Molino river reach of 30.8 km distance within the same basin where the wave travel time is 3h and drainage area is about 1135 km(2). Inflow hydrographs were generated and compared against the observed ones. The model simulation of inflow hydrographs were satisfactory.Article Citation - WoS: 27Citation - Scopus: 29Fuzzy, Ann, and Regression Models To Predict Longitudinal Dispersion Coefficient in Natural Streams(IWA Publishing, 2006) Tayfur, GökmenThis study developed fuzzy, ANN, and regression-based models to predict longitudinal dispersion coefficient in natural streams from flow discharge data. 92 sets of field data were employed to calibrate and validate the models. 63 sets of data were used for the calibration while the remaining data were used for the validation of the models. The model-prediction results revealed the superiority of the developed models over the existing equations. The developed models predicted the measured data satisfactorily with minimum errors and maximum accuracy rates. The three models had comparable performances although the fuzzy model had the highest accuracy rate (79%) and lowest mean relative error (0.85).Article Citation - WoS: 1Citation - Scopus: 1Generalized Regression Neural Network and Empirical Models To Predict the Strength of Gypsum Pastes Containing Fly Ash and Blast Furnace Slag(Springer Verlag, 2020) Erdem, Tahir Kemal; Cengiz, Okan; Tayfur, GökmenGypsum is widely used in constructions owing to its easy application, zero shrinkage, and excellent fire resistance. Several parameters can affect the properties of gypsum pastes. To study the strength of the gypsum pastes experimentally by trying all these parameters is time-consuming and costly. Therefore, artificial intelligence methods can be very useful to predict the paste strength, which, in turn, can reduce the number of trial batches. Based on experimental data, the generalized regression neural network (GRNN) and empirical models were developed to predict strength of gypsum pastes containing fly ash (FA) and blast furnace slag (BFS). Gypsum content, pozzolan content, curing temperature, curing duration, and testing age constituted the input variables of the models while the paste strength was the target output. The trained and tested GRNN model was found to be successful in predicting strength. Sensitivity analysis by the GRNN model revealed that the curing duration and temperature were important sensitive parameters. In addition to the GRNN model, empirical models were proposed for the strength prediction. The same input variables formed the input vectors of the empirical models. The same dataset used for the calibration of the GRNN model was employed to establish the empirical models by employing genetic algorithm (GA) method. The empirical models were successfully validated. The GRNN and GA_based empirical models were also tested against the multi-linear regression (MLR) and multi-nonlinear regression (MNLR) models. The results showed the outperformance of the GRNN and the GA_based empirical models over the others.