Sürdürülebilir Kalkınma Amaçları
Yazarlar
- Kankal, Murat(3)
- Akpinar, Adem(2)
- Duran, Celal(2)
- Kömürcü, Murat Ihsan(2)
- Ozdes, Duygu(2)
- Akçay, Serhatcan Berk(1)
- Alemdag, Selcuk(1)
- Altas, Ismail H.(1)
- Bacak, Aykut(1)
- Bayram, Adem(1)
- Başoğlu, Mustafa Engin(1)
- Beder, Murat(1)
- Bektas, Sibel Cevik(1)
- Benli, Büsra(1)
- Cakmak, Recep(1)
- Camoglu, Aslihan Yilmaz(1)
- Celen, Ali(1)
- Dalkiliç, Ahmet Selim(1)
- Es, Hüseyin Avni(1)
- Gurocak, Zulfu(1)
- Güneroğlu, Aziz(1)
- Hamzaçebi, Coşkun(1)
- Karabi̇Naoğlu, Murat Salim(1)
- Karatas, Seyda Merve(1)
- Kavuncuoglu, Erhan(1)
- Kavuncuoglu, Hatice(1)
- Kazdaloğlu, Abdülvehhab(1)
- Kose, Oguz(1)
- Oktay, Tugrul(1)
- Ozeken, Sengul Tugba(1)
- Ozsahin, Talat Sukru(1)
- Sagdic, Osman(1)
- Saka, Fatih(1)
- Seker, Serdal(1)
- Sezer, Cihan(1)
- Sezer, Sukru(1)
- Solanki, Pranshoo(1)
- Tilki, Nida(1)
- Varol, Temel(1)
- Yalcin, Hasan(1)
Yayınlar
- MakaleAçık Erişim2024Prediction of effect of fabrication parameters on the properties of B4C ceramic particle reinforced AA2024 matrix nanocomposites using neural networks
Varol, Temel; Çanakçı, Aykut; Özşahin, Şükrü; Beder, Murat; Akçay, Serhatcan Berk
In the present study, a prediction model was created to predict the impact of B4C content, milling time and matrix size on the hardness and density of AA2024-B4C nanocomposites produced by means of flake powder metallurgy. The process involved synthesizing AA2024-B4C nanocomposite powders by mechanical milling followed by preparing nanocomposites by hot pressing. The matrix size of the AA2024 alloy powders, the milling time, and the B4C content were selected as the input parameters. On the other hand, the density and the hardness of the AA2024-B4C nanocomposites were designed as the output parameters. The results showed that the sintered density is significantly improved by decreasing matrix size. It was found that density of AA2024–3 wt%B4C nanocomposites is lower than AA2024–1 wt%B4C nanocomposites. However, the hardness of AA2024–3 wt%B4C nanocomposites is better than AA2024–1 wt%B4C nanocomposites. The AA2024-B4C nanocomposite prediction model could predict hardness and density with a mean absolute percentage error (MAPE) of 0.32% and 1.38%. The results show that the ANN model is a useful tool for the matrix size effect prediction, milling time and B4C content on the density and hardness of AA2024-B4C nanocomposites prepared by flake powder metallurgy.
- MakaleAçık Erişim2024Investigation on the heat transfer and pressure loss of flow boiling in smooth and microfin tubes using machine learning methods
Sezer, Sukru; Sezer, Cihan; Celen, Ali; Bacak, Aykut; Dalkiliç, Ahmet Selim
The estimation of heat transfer coefficients (HTC) and pressure drop (Delta P) in flow boiling processes is essential for the effective design and operation of refrigeration systems. In this study, the artificial neural network (ANN), locally weighted regression (LWR), and gradient boosted machine (GBM) methods are employed to predict the boiling heat transfer coefficient (HTC) and pressure drop (Delta P\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}Δ P\end{document}) in flow boiling of R134a. The study focuses on horizontally positioned both straight and microfin tubes. The ANN, LWR, and GBM methodologies are utilized to ascertain the parameters of boiling HTC and Delta P as outputs. These parameters are determined by considering the mass flux, saturation pressure, heat flux, vapor quality, Reynolds number, Lockhart-Martinelli parameter, Froud number, Weber number, and Bond number as inputs. The training dataset is partitioned into 5 sections for the purpose of hyperparameter tweaking for each model. Out of these sections, 4 parts, consisting of approximately 111 samples, are utilized for training, while 1 part, including around 27 samples, is allocated for validation. The optimal hyperparameters are determined by calculating the average R2 score over the 5 validation sets. Using raw measurements, HTC and Delta P are successfully modeled using a relatively much smaller dataset of 174 measurements, with 82.4% R2 score and 0.7% weighted average relative deviation for HTC, and 88.9% R2 score and 4.1% weighted average relative deviation for Delta P across multiple tube types, achieved by LWR algorithm. Model performances are validated with an extrapolation test and found to be consistent with traditional train-validation-test sampling scheme with 75.9% R2 score and -6.2% weighted average relative deviation for HTC, and 89.3% R2 score and -3.9% weighted average relative deviation for Delta P, showing the consistency of the hypotheses created by a hybrid of parametric and nonparametric model families even outside the observed measurement range for multiple tube types. Local weighted regression models are the most performant, especially for limited data availability. However, calculated measurements increase error rates, suggesting that HTC and Delta P models work best with raw measurements.
- MakaleAçık Erişim2023Enhancement of adsorptive removal efficiency of an anionic dye from aqueous solutions using carboxylic acid-modified mulberry leaves: Artificial neural network modeling, isotherm, and kinetics evaluation
Duran, Celal; Ozeken, Sengul Tugba; Camoglu, Aslihan Yilmaz; Ozdes, Duygu
Natural mulberry leaves and carboxylic acid-modified mulberry (Morus alba L.) leaves were used for the first time to scrutinize the effects of modification on the retention efficiency of an anionic dye (Remazol Brilliant Blue R (RBBR)) from aqueous solutions to suggest an economical and promising adsorbent for the treatment of dye-contaminated water. The characterization of the adsorbents was accomplished through common techniques including SEM, FTIR, and pHpzc determination. Several parameters studied in batch experiments pointed out that the initial pH of 2.0 and the contact time of 240 min were optimum conditions for all the developed RBBR uptake processes. An artificial neural network (ANN) model was applied to formulate a forecast model for the uptake efficiency of RBBR. The experimental data were assessed by different kinetic and isotherm models to explain the mechanism of the developed processes in more detail. Maximum monolayer adsorption capacities of natural mulberry leaves and acetic acid-, citric acid-, and oxalic acid-modified mulberry leaves were determined as 64.5, 95.2, 84.8, and 91.7 mg g1, respectively, by the Langmuir isotherm model. These results demonstrated that the modification with carboxylic acids significantly increases the anionic dye adsorption capacity of the mulberry leaves.
- MakaleAçık Erişim2022Comparison of deep learning and regression-based MPPT algorithms in PV systems
Karabi̇Naoğlu, Murat Salim; Çakir, Bekir; Başoğlu, Mustafa Engin; Kazdaloğlu, Abdülvehhab; Güneroğlu, Aziz
Solar energy systems (SES) and photovoltaic (PV) modules should be operated at the maximum power point (MPP) to achieve the highest efficiency in the energy generation processes. Maximum power point tracking (MPPT) applications using conventional methods may not be able to follow the global MPP (GMPP) of the PV system under changing atmospheric conditions and they could oscillate around the local MPP. In this study, a machine learning and deep learning (DL) based long short-term memory (LSTM) model is proposed as an innovative solution for MPPT. Contrary to the traditional MPPT applications using current and voltage sensors, the output resistance of the PV module estimation was made by using environmental parameters (such as temperature and radiation) and artificial intelligence algorithms in this study.The LSTM model was compared with artificial neural networks (ANN) and regression methods regarding mean square error (MSE), root mean square error(RMSE) and mean absolute error (MAE) parameters. It has been determined that the LSTM model has a better performance and could more successfully follow MPP compared to the other methods. Finally, after the comparison with the ANN method, it is proved that LSTM gives 37%, 21%, and 31% more successful MSE, RMSE, and MAE results, respectively.
- MakaleAçık Erişim2022Insights into brewed tea waste as a green and low-priced adsorbent for solid-phase extraction of Cd(II) ions: isotherm, kinetic, and artificial neural network approach
Ozdes, Duygu; Tilki, Nida; Seker, Serdal; Duran, Celal
In the present research, brewed tea waste (BTW) was utilized as a green, low-priced, and abundant adsorbent for separation/preconcentration of Cd(II) ions through solid-phase extraction method from water and foods for the first time. BTW was applied as a natural adsorbent, without using any chelating agent to bind Cd(II) ions or any chemical reagent for its modification. A three-layer artificial neural network model using backpropagation algorithm was utilized to explicate a prediction model for the extraction performance of Cd(II) ions by selecting the input parameters as solution pH, quantity of BTW, sample volume, eluent concentration and volume, and equilibrium time for desorption. The preconcentration factor, relative standard deviation, and detection limit were attained as 100, 3.03%, and 0.56 µg L−1, respectively. It was decided that the Langmuir isotherm model is acceptable to characterize the retention of Cd(II) ions on BTW. This result pointed out that the active binding sites on the BTW surface are homogeneously distributed. Adsorption capacity of BTW was achieved as 41.5 mg g−1 which is higher than several expensive and difficult-to-prepare adsorbents. Adsorption kinetics was elucidated by pseudo-second order kinetic model. After confirmed the accuracy of the method with spike/recovery studies, it was employed for Cd(II) determination in water (stream and sea water) and food (eggplant, lettuce, parsley, apple, and apricot) samples with high accuracy. The inferences of the study proved that the BTW offers a magnificent application prospect in the extraction of Cd(II) ions.
- MakaleAçık Erişim2021Quadrotor Flight System Design using Collective and Differential Morphing with SPSA and ANN
Kose, Oguz; Oktay, Tugrul
Quadrotor modeling has been done with collective and differential morphing. Quadrotor initial state and morphing states are drawn in the Solidworks program. Newton-Euler approximation was used for quadrotor modeling. The mass and moment of inertia values required for modeling and simulation were obtained from the Solidworks program. Matlab / Simulink environment and state-space model approaches are used for simulations. A simultaneous perturbation stochastic approximation (SPSA) algorithm was used to determine the quadrotor morphing rates. If the morphing state obtained by SPSA is not included in the values obtained from the drawings, here it is provided to find the moments of inertia with the method based on learning by using the data obtained with the Artificial Neural Network(ANN). Proportional Integral Derivative (PID) is used as the quadrotor control algorithm. PID coefficients are also determined by SPSA.
- MakaleYalnız künye2019Forecasting of Turkey’s monthly electricity demand by seasonal artificial neural network
Hamzaçebi, Coşkun; Es, Hüseyin Avni; Çakmak, Recep
Electricity is one of the most important end-user energy types in today’s world and has an effective role in development of societies and economies. Stability of electricity supply is provided by matching of generated and consumed electricity amount during the all-day. So, electricity consumption forecasting is an essential issue for electric utilities. In this study, the monthly electricity demand of Turkey has been predicted. To model the effects of seasonality and trend, four different ANN models have been developed and selected the superior one. In addition, the selected ANN model has been compared with SARIMA model in order to increase the acceptability and reliability of the ANN model. The monthly electricity demand of Turkey has been predicted between 2015 and 2018 via the ANN model that can make successful and high-accuracy predictions according to the performance measures. The forecasting values will help in determining the medium-term and stable energy policies.
- BildiriYalnız künye2019Design of a MATLAB GUI for Day Ahead Forecasting of PV Panel Power
Bektas, Sibel Cevik; Cakmak, Recep; Altas, Ismail H.
This paper introduces design and utilization of a MATLAB graphical user interface (GUI) platform to forecast day ahead hourly Photovoltaic (PV) panel power by using forecasted day ahead global horizontal solar radiation. This platform, which would be a useful and practical tool for researchers who are study on energy management, planning and forecasting studies, consists of three parts. The first part predicts day ahead global horizontal solar radiation by trained artificial neural network (ANN) model. The second part performs the calculation of solar irradiation on the module, which has a tilted angle according to forecasting solar radiation data. The last part estimates the power generated by PV panel according to the calculated incident radiation. In this study, the solar irradiation forecasting, the calculation of incident radiation to tilted module and prediction of the solar power to be generated by the PV module are done. The designed GUI has been utilized to forecast solar radiation and PV output power for test days. Then obtained forecasted data have been evaluated with mean absolute percentage error (MAPE) and correlation coefficient (R).
- MakaleYalnız künye2018Prediction of the antimicrobial activity of walnut (Juglans regia L.) kernel aqueous extracts using artificial neural network and multiple linear regression
Kavuncuoglu, Hatice; Kavuncuoglu, Erhan; Karatas, Seyda Merve; Benli, Büsra; Sagdic, Osman; Yalcin, Hasan
The mathematical model was established to determine the diameter of inhibition zone of the walnut extract on the twelve bacterial species. Type of extraction, concentration, and pathogens were taken as input variables. Two models were used with the aim of designing this system. One of them was developed with artificial neural networks (ANN), and the other was formed with multiple linear regression (MLR). Four common training algorithms were used. Levenberg–Marquardt (LM), Bayesian regulation (BR), scaled conjugate gradient (SCG) and resilient back propagation (RP) were investigated, and the algorithms were compared. Root mean squared error and correlation coefficient were evaluated as performance criteria. When these criteria were analyzed, ANN showed high prediction performance, while MLR showed low prediction performance. As a result, it is seen that when the different input values are provided to the system developed with ANN, the most accurate inhibition zone (IZ) estimates were obtained. The results of this study could offer new perspectives, particularly in the field of microbiology, because these could be applied to other type of extraction, concentrations, and pathogens, without resorting to experiments.
- MakaleYalnız künye2012Artificial neural networks for estimation of temporal rate coefficient of equilibrium bar volume
Kankal, Murat; Kömürcü, Murat Ihsan; Yüksek, Ömer; Akpinar, Adem
Present study consists the growth of a bar profile caused by cross-shore sediment transport. This is especially on growth of bar volume (V) toward equilibrium bar volume (Veq). Three analysis methods being a power and linear regression analysis (PRA and LRA) and an Artificial Neural Network (ANN) analysis were performed to determine empirical temporal rate coefficient (α). Forty-two experimental data were used for training set and the rest of the experimental data were used for testing set in the ANN analysis. As the results of analyses, the smallest average relative and root mean square error (RMSE) computed for the ANN methods are 7.578% and 0.029, respectively. It has been obtained that the ANN analysis, which is used for determination of α coefficient, gives reasonable results. Finally, bar volumes were calculated by means of computed α values and compared with the results of experimental data.
- MakaleAçık Erişim2012New considerations for empirical estimation of tensile strength of rocks
Gurocak, Zulfu; Solanki, Pranshoo; Alemdag, Selcuk; Zaman, Musharraf M.
In this paper, a combined laboratory and modeling study was conducted to develop a database for predicting tensile strength of rocks. Six hundred eighty six rock samples from 24 different sites throughout eastern Turkey were collected and tested for the development of this database and evaluation of models. A total of 512 samples were used for developing the models and the remaining 174 samples were used as control dataset. The material parameters selected in the development of the models include tensile strength (σ t), point load index (Is (50)), Schmidt rebound number (N) and unit weight (γ). A total of four models, two regression models, namely, simple linear regression and multiple regression, and two feed forward-type artificial neural network (ANN) models, namely, radial basis function network (RBFN) and multi-layer perceptron network (MLPN) are developed. A commercial software, Statistica 8.0, is used to develop these models. The strengths and weaknesses of the developed models were assessed by comparing the predicted σ t values with the experimental values with respect to the R 2 values. Overall, the MLPN model was found to be the best model for the present development and evaluation datasets. As a result of these analyses, an equation was suggested based on ANN model to estimate the tensile strength of rocks.
- MakaleYalnız künye2011Estimation of the carbon to nitrogen (C:N) ratio in compostable solid waste using artificial neural networks
Bayram, Adem; Kankal, Murat; Ozsahin, Talat Sukru; Saka, Fatih
Organic waste constitutes a majority of all municipal solid waste (MSW), a fact which yields some unfavorable results at open dumps, sanitary landfills, and incineration plants. As part of an integrated solid waste management strategy, composting could be applied to mixed collected MSWs or separately collected leaves, food, and yard wastes. The factor most crucial to successful composting is the carbon to nitrogen (C:N) ratio of the waste. This study employs two predictive models to estimate the C:N ratio of compostable MSW, an artificial neural network (ANN) and multiple linear regression (MLR). These models are based on 52 solid waste samples taken from the MSW open dumping area in Gümüşhane Province, Turkey. To estimate the C:N ratio, seven predictive variables were adopted. The proportions of food and yard (F&Y), and ash and scoria (A&S) waste; the moisture content (MC), the fixed carbon (FC) content, the total amount of organic matter (TOM), high calorific value (HCV), and pH. Fortytwo of the samples were used for training, and the remaining ten were used to test the models. The average relative error attained by the best ANN was 6.376%, while that attained by the MLR model was 11.002%. The effects of TOM content, F&Y percentage, and A&S percentage on the C:N ratio were investigated by running the ANN model for a range of input variables.
- MakaleYalnız künye2011Modeling and forecasting of Turkey's energy consumption using socio-economic and demographic variables
Kankal, Murat; Akpinar, Adem; Kömürcü, Murat Ihsan; Özşahin, Talat Şükrü
This study deals with the modeling of the energy consumption in Turkey in order to forecast future projections based on socio-economic and demographic variables (gross domestic product-GDP, population, import and export amounts, and employment) using artificial neural network (ANN) and regression analyses. For this purpose, four diverse models including different indicators were used in the analyses. As the result of the analyses, this research proposes Model 2 as a suitable ANN model (having four independent variables being GDP, population, the amount of import and export) to efficiently estimate the energy consumption for Turkey. The proposed model predicted the energy consumption better than the regression models and the other three ANN models. Thus, the future energy consumption of Turkey is calculated by means of this model under different scenarios. The predicted forecast results by ANN were compared with the official forecasts. Finally, it was concluded that all the scenarios that were analyzed gave lower estimates of the energy consumption than the MENR projections and these scenarios also showed that the future energy consumption of Turkey would vary between 117.0 and 175.4. Mtoe in 2014.