DSpace Arşivi :: by Yazar "Colak, Andac Batur" değerine göre listeleniyor

Yazar "Colak, Andac Batur" seçeneğine göre listele

Listeleniyor 1 - 20 / 50

A COMPREHENSIVE AND COMPARATIVE EXPERIMENTAL ANALYSIS ON THERMAL CONDUCTIVITY OF TiO2-CaCO3/WATER HYBRID NANOFLUID: PROPOSING NEW CORRELATION AND ARTIFICIAL NEURAL NETWORK OPTIMIZATION
(Begell House Inc, 2021) Ocal, Sultan; Gokcek, Murat; Colak, Andac Batur; Korkanc, Mustafa
In this study, the thermal conductivity of TiO2-CaCO3/water hybrid nanofluid, which was prepared with five different concentrations and two-step method, was experimentally investigated. Thermal conductivity measurements were made using the KD2 Pro device at a temperature range from 10 degrees C to 60 degrees C. Using experimental data, a mathematical correlation and an artificial neural network model was developed in order to predict thermal conductivity depending on concentration and temperature. In the feed-forward back-propagation artificial neural network with 10 neurons in its hidden layer, the multilayer perceptron model was preferred. While the value of the coefficient of determination R for the proposed new mathematical correlation was 0.9999, it was obtained as 0.99913 for the artificial neural network model. The average error rate was calculated as 0.005% for the mathematical model and -0.02% for the artificial neural network.
A novel comparative analysis between the experimental and numeric methods on viscosity of zirconium oxide nano fluid: Developing optimal artificial neural network and new mathematical model
(Elsevier, 2021) Colak, Andac Batur
In this study, the viscosity of five different ZrO2/Water nanofluids of 0.0125%, 0.025%, 0.05%, 0.1% and 0.2% prepared by the two-step method were experimentally investigated. Using the experimental data obtained, a multi-layer perceptron feed-forward back-propagation artificial neural network and a new mathematical correlation have been developed in order to predict the viscosity of ZrO2/Water nanofluid. Experimental results revealed that viscosity decreases with increasing temperature and increases with increasing concentration. The outputs obtained from the developed artificial neural network and the new correlation were compared with the experimental results and analyzed. The results show that the developed artificial neural network can predict the viscosity of ZrO2/Water nanofluid with an average error rate of -0.11%. The new mathematical model developed has been able to calculate the viscosity of ZrO2/Water nanofluid with an error rate of -0.74%. (C) 2020 Elsevier B.V. All rights reserved.
A novel comparative investigation of the effect of the number of neurons on the predictive performance of the artificial neural network: An experimental study on the thermal conductivity of ZrO2 nanofluid
(Wiley-Hindawi, 2021) Colak, Andac Batur
In this study, the effect of the number of neurons on the predictive performance of artificial neural networks (ANN) has been investigated using experimental data. For this purpose, 6 different ANN have been developed by using a total of 60 experimental data of ZrO2/water nanofluid obtained from the literature. In ANN developed with the number of 5, 10, 15, 20, 25, and 30 neurons, all other parameters have been kept constant, and the effect of only the number of neurons on the prediction performance has been investigated. The performance of each ANN has been calculated separately and then their performance has been analyzed by comparing them with each other. As a consequence of the study, it has been seen that the model with the most ideal predictive performance has been developed with 5 neurons with an average error rate of 0.001%, and the highest margin of error the model has been developed with 15 neurons and had an error rate of 0.026%. In the light of the obtained data, it has been concluded that ANN are generally high performance predictive tools, and it is not possible to reach a standard correlation to regulate the number of neurons to be used in the optimization of ANN.
A NOVEL MACHINE LEARNING STUDY: MAXIMIZING THE EFFICIENCY OF PARABOLIC TROUGH SOLAR COLLECTORS WITH ENGINE OIL-BASED COPPER AND SILVER NANOFLUIDS
(Begell House Inc, 2024) Colak, Andac Batur; Bayrak, Mustafa
Estimating the heat transfer parameters of parabolic trough solar collectors with machine learning is crucial for improving the efficiency and performance of these renewable energy systems, optimizing their design and operation, and reducing costs while increasing the use of solar energy as a sustainable power source. In this study, the heat transfer characteristics of two different nanofluids flowing through the porous media in a straight plane underneath thermal jump conditions were investigated by machine learning methods. For the flow in the parabolic trough solar collector, two different nanofluids obtained from silver- and copper-based motor oil are considered. Flow characteristics were obtained by nonlinear surface tension, thermal radiation, and Cattaneo-Christov heat flow, which was used to calculate the heat flow in the thermal boundary layer. A neural network structure was established to estimate the skin friction and Nusselt number determined for the analysis of the flow characteristic. The data used in the multilayer neural network, which was developed using a total of 30 data sets, were divided into three groups as training, validation, and testing. In the input layer of the network model with 15 neurons in the hidden layer, 10 parameters were defined and four different results were obtained for two different nanofluids in the output layer. The prediction performance of the established neural network model has been comprehensively studied by means of several performance parameters. The study findings presented that the established artificial neural network can predict the heat transfer characteristics of two different nanofluids obtained from silver- and copper-based motor oil with deviation rates less than 0.06%.
A NUMERICAL STUDY AIMED AT FINDING OPTIMAL ARTIFICIAL NEURAL NETWORK MODEL COVERING EXPERIMENTALLY OBTAINED HEAT TRANSFER CHARACTERISTICS OF HYDRONIC UNDERFLOOR RADIANT HEATING SYSTEMS RUNNING VARIOUS NANOFLUIDS
(Begell House Inc, 2022) Colak, Andac Batur; Karakoyun, Yakup; Acikgoz, Ozgen; Yumurtaci, Zehra; Dalkilic, Ahmet Selim
In this paper, three unique artificial neural network models have been developed for three different working fluid cases to predict the radiative, convective, and total heat transfer coefficients over the floor surface of radiant floor heating system in a real-size room. Pure water, multiwall carbon nanotube with 0.7 vol.% and 0.07 vol.% contents, and aluminium oxide with 1.26 vol.% content are the operating fluids having inlet temperatures ranging from 30 degrees C to 60 degrees C, while the mass flow rates are 0.056, 0.09, and 0.125 kg/s. The performances of multilayer perceptron networks with the Levenberg-Marquardt, Bayesian regularization, and scaled conjugate gradient as training algorithms and different neuron numbers have been developed and the Levenberg-Marquardt algorithm, having the highest prediction performance with 99% accuracy, is selected as a result of detailed computational numerical analyses. This study can be considered as a pioneer artificial neural network one on the floor heating systems having nanofluids.
An experimental and new study on thermal conductivity and zeta potential of Fe3O4/water nanofluid: Machine learning modeling and proposing a new correlation
(Elsevier, 2023) Sahin, Fevzi; Genc, Omer; Gokcek, Murat; Colak, Andac Batur
It is important to predict the thermophysical properties of nanofluids, which have higher heat transfer perfor-mance compared to the base fluid, without the need for experimental studies. In this study, two different artificial neural networks were created to predict the thermal conductivity and zeta potential of Fe3O4/water nanofluid. The thermal conductivity and zeta potential of the Fe3O4/water nanofluid prepared at three different concen-trations were experimentally measured. An innovative mathematical correlation is proposed to calculate thermal conductivity based on temperature and concentration using the obtained experimental data. Considering that the correlations in the literature can generally be calculated according to concentration, the novelty of the proposed model stands out. The calculated values for thermal conductivity and zeta potential of the created artificial neural network and the new mathematical correlation were compared with the results of the experiments. In addition, a comprehensive performance analysis was made by calculating different performance parameters. The R values of the neural network models were above 0.99 and mean squared error values were obtained as 1.47E-05 and 1.58E-06, respectively. In addition, the mean deviation values calculated for the thermal conductivity of the network model were 0.03%, while it was 0.05% for the new mathematical correlation. The study results showed that ANN models can predict the thermal conductivity and zeta potential of Fe3O4/water nanofluid with high accuracy. The proposed new mathematical correlation was also found to have higher error rates compared to the ANN model, although it was able to calculate thermal conductivity values with high accuracy.
AN EXPERIMENTAL EVALUATION OF WORKABILITY AND BLEEDING BEHAVIORS OF ULTRA-SONICATED NANO ZINC OXIDE (n-ZnO) DOPED CEMENT PASTES INCORPORATED WITH FLY ASH
(Begell House Inc, 2024) Celik, Fatih; Colak, Andac Batur; Yildiz, Oguzhan; Bozkir, Samet Mufit
In this study, the workability and bleeding behaviors of ultra-sonicated nano zinc oxide (n-ZnO) doped cement pastes incorporated with fly ash have been experimentally investigated. Therefore, the effects of nano zinc oxide (n-ZnO) additions at different amounts by mass (0.0, 0.3, 0.6, 0.9, 1.2, and 1.5%) on the bleeding and the workability properties (mars cone flow time, mini slump spread diameter, and plate cohesion) of cement -based grouts incorporated with fly ash (FA) as mineral additive at different constitutes (0% -for control purpose, 5, 10, 15, 20, 25, and 30%) were investigated. The use of FA as a mineral additive in grout samples resulted in improvements in the workability behavior of the grout samples as expected. Increase amount of n-ZnO in the grout mixtures has made mini slump flow diameter of the samples noticeably decrease. Although certain changes seem to have been observed, it has been understood that the increase in the amount of n-ZnO in the injection matrix generally does not change the Marsh cone flow time of mineral -added cement -based grouts. Remarkable increases in plate cohesion values were measured because of the increase in the content of nano zinc oxide for all mixtures. At the same time, just like the FA effect, bleeding values tend to decrease due to the increase in the amount of nano zinc oxide in grout mixes. Moreover, the results obtained showed that the artificial neural network model can make predictions with very high accuracy.
An Experimental Investigation on Workability and Bleeding Behaviors of Cement Pastes Doped with Nano Titanium Oxide (n-TiO2) Nanoparticles and Fly Ash
(Tech Science Press, 2023) Celik, Fatih; Yildiz, Oguzhan; Colak, Andac Batur; Bozkir, Samet Mufit
In this study, the workability of cement-based grouts containing n-TiO2 nanoparticles and fly ash has been investigated experimentally. Several characteristic quantities (including, but not limited to, the marsh cone flow time, the mini slump spreading diameter and the plate cohesion meter value) have been measured for different percentages of these additives. The use of fly ash as a mineral additive has been found to result in improvements in terms of workability behavior as expected. Moreover, if nano titanium oxide is also used, an improvement can be obtained regarding the bleeding values for the cement-based grout mixes. Using such experimental data, a multi-layer perceptron artificial neural network model has been developed (5 neurons in the hidden layer of the network model have been developed using a total of 42 experimental data). 70% of the data employed in this model have been used for training, 15% for validation and 15% for the test phase. The results demonstrate that the artificial neural network model can predict Marsh cone flow time, mini slump spreading diameter and plate cohesion meter values with an average error of 0.15%.
An Experimental Investigation on Workability and Bleeding Features
(Amer Concrete Inst, 2022) Celik, Fatih; Colak, Andac Batur; Yildiz, Oguzhan; Bozkir, Samet Mufit
In this experimental study, the workability and bleeding properties of cement-based grout mixtures combined with fly ash (FA) and colloidal nanopowder (n-Al2O3) were investigated, and some prediction models were developed with an artificial neural network (ANN). Marsh cone flow time, mini-slump spreading diameter, and Lombardi plate cohesion of the grout samples were measured based on the workability test. Test results showed that the use of FA as mineral additive in the grout samples positively contributed to an increase of the fluidity of the grout samples as expected. Considerable effects were observed on workability features of grout mixtures with the addition of nano alumina because of having a large specific surface area. In addition, the use of nano alumina together with FA in grout mixtures contributes to the stability of these mixtures by looking at changes in bleeding values. Using the experimental data obtained, an ANN model was developed to predict the values of Marsh cone flow time, mini-slump spreading diameter, and plate cohesion. The developed ANN model can predict mini-slump spreading diameter with an error rate of -0.04%, Marsh cone flow time value with an error rate of -0.23%, and plate cohesion value with an error rate of -1.07%.
An Experimental Study on Artificial Intelligence-Based Prediction of Capacitance-Voltage Parameters of Polymer-Interface 6H-SiC/MEH-PPV/Al Schottky Diodes
(Wiley-V C H Verlag Gmbh, 2022) Guzel, Tamer; Colak, Andac Batur
Herein, an artificial neural network (ANN) model has been developed to predict the capacitance values of the polymer-interface 6H-SiC/MEH-PPV/Al Schottky diode depending on the frequency. In the training of the feed-forward back-propagation network model with five neurons in its hidden layer, 480 experimental data have been used. Of these, 70% of the data used in the development of the multilayer perceptron network has been used for network training, 15% for validation, and 15% for the test phase. The predictive performance of the network model has been analyzed by comparing the predicted values obtained from the ANN with the experimental data. For the developed ANN, the mean square error value is 4.34E-06, the R-value is 0.99728, and the average margin of deviation value is 0.03%.
An experimental study on determination of the shottky diode current-voltage characteristic depending on temperature with artificial neural network
(Elsevier, 2021) Colak, Andac Batur; Guzel, Tamer; Yildiz, Oguzhan; Ozer, Metin
Shottky diodes are one of the important components of electronic systems. Therefore, it is very important to determine the parameters of the diodes according to the area in which they will be used. One of the most important of these parameters is the current-voltage characteristic of the diode. In this study, firstly, current values of the Schottky diode in the voltage range of -2 V to +3 V are experimentally measured in the temperature range of 100?300 K. In order to estimate the current-voltage characteristic of Shottky diode at different temperatures, a multi-layer perceptron, a feed-forward back-propagation artificial neural network was developed using 362 experimental data obtained. In the artificial neural network where temperature (T) and voltage (V) values are selected as input variables and the hidden layer has 15 neurons, the current (I) value is obtained as output. The results obtained from the artificial neural network have been found to be in good agreement with the experimental data of the Schottky diode.
An experimental study on the comparative analysis of the effect of the number of data on the error rates of artificial neural networks
(Wiley, 2021) Colak, Andac Batur
In this study, the effect of the amount of data used in the design of artificial neural networks (ANNs) on the predictive accuracy of ANNs was investigated. Five different ANNs were designed using the experimentally measured specific heat data of the Al2O3/water nanofluid prepared at volumetric concentrations of 0.0125, 0.025, 0.05, 0.1 and 0.2 using the Al(2)O(3)nanoparticle. The developed ANN is a multi-layer perceptron, feedforward and backpropagation model. In each ANN with 15 neurons in the hidden layer, the volumetric concentration (phi) and temperature (T) values were nominated as input layer factors and the specific heat value was estimated as the output value. With the aim of survey the effect of the amount of data on the predicted results of the ANN, a different amount of datasets were used in each developed ANN. In this context, in total 260 data were used in the Model 1 ANN. Subsequently, the total amount of data was reduced by 20% in each developed neural network and 55 data were used in the ANN named Model#5. The results obtained show that ANNs are highly talented of predicting the specific heat values of Al2O3/water nanofluid. However, in the comparisons, it was evaluated that the amount of data used had a share on the prediction performance of the ANN and that the decrease in the amount of data with the prediction performance of the ANN decreased.
Analysing of nano-silica usage with fly ash for grouts with artificial neural network models
(Emerald Group Publishing Ltd, 2023) Celik, Fatih; Yildiz, Oguzhan; Colak, Andac Batur; Bozkir, Samet Mufit
When grout is used to penetrate voids and cracks in soils and rock layers, easy and effective pumping of the grouts is vital, especially for grouting works during geotechnical improvements. For this reason, improving the rheological parameters of cement-based grouts and increasing their fluidity are important for effective grouting injection. In this study, an experimental investigation and analysis using artificial neural network (ANN) models were used to discover how nano silica (n-SiO2) together with fly ash affects the rheological behaviour of cement-based grouts. The effects of nano silica (n-SiO2) additions at different contents by mass (0.0%, 0.3%, 0.6%, 0.9%, 1.2% and 1.5%) on the plastic viscosity and yield stress values of cement-based grouts incorporating fly ash as a mineral additive at different amounts (0% - as a control, 5%, 10%, 15%, 20%, 25% and 30%) were investigated. Using the experimental data obtained, a feed-forward (FF) back-propagation (BP) multi-layer perceptron (MLP) artificial neural network (ANN) was developed to predict the plastic viscosity and yield stress of cement-based grouts with nano silica nanoparticle additives. The ANN model developed can predict the plastic viscosity and yield stress values of cement-based grouts containing nano silica nanoparticle-doped fly ash with high accuracy.
Artificial Intelligence Approach in Predicting the Effect of Elevated Temperature on the Mechanical Properties of PET Aggregate Mortars: An Experimental Study
(Springer Heidelberg, 2021) Colak, Andac Batur; Akcaozoglu, Kubilay; Akcaozoglu, Semiha; Beller, Gulhan
In this study, the effect of high temperature on the flexural and compressive strength of mortars containing waste PET aggregates was investigated experimentally. The mortar samples prepared in 5 different concentrations with a total of 2.5%, 5%, 10%, 20% and 30% PET aggregate substitution were heated up to 100, 150, 200, 250, 300 and 400 degrees C. After waiting for 1, 2 and 3 h at these temperatures, flexural and compressive strength tests were performed. It was observed that flexural strength and compressive strength values decreased with increasing temperature and PET aggregate amounts in all mixtures. An artificial neural network was designed to estimate flexural and compressive strength values using experimental data. It has been observed that the developed artificial neural network can predict flexural and compressive strengths with an average error of - 0.51%.
Artificial intelligence approach on predicting current values of polymer interface Schottky diode based on temperature and voltage: An experimental study
(Academic Press Ltd- Elsevier Science Ltd, 2021) Guzel, Tamer; Colak, Andac Batur
In this study, an artificial neural network model has been developed to predict the current values of a 6H?SiC/MEH-PPV Schottky diode with polymer-interface, depending on temperature and voltage. In the training of the multi-layer perceptron network model with 13 neurons in its hidden layer, the experimentally measured current values between 100 and 250 K temperature and -3V to + 3V voltage range have been used. In the input layer of the model developed with a total of 244 experimental data, temperature, and voltage values have been defined and current values were obtained in the output layer. The mean square error value of the artificial neural network is 1.63E-08 and the R-value is 0.99999. The developed model has been able to predict the current values of the polymer-interfaced 6H?SiC/MEH-PPV Schottky diode with an average error rate of -0.15% depending on temperature and voltage, with high accuracy.
Artificial neural network approach for investigating the impact of convector design parameters on the heat transfer and total weight of panel radiators
(Elsevier France-Editions Scientifiques Medicales Elsevier, 2023) Calisir, Tamer; Colak, Andac Batur; Aydin, Devrim; Dalkilic, Ahmet Selim; Baskaya, Senol
The difficulty of the production stages of panel radiators used for heating purposes reveals the importance of determining the heat transfer performance and panel radiator weight values, which are determined depending on the design parameters. In the present work, an artificial neural network model is proposed for predicting the heat transfer and weight values of a panel radiator as outputs depending on the design parameters of convectors. In the multilayer network model developed with 78 numerically obtained data sets, 8 different design parameters were defined as input parameters and heat transfer and in the output layer panel weight values were obtained. The design parameters of the convectors, in other words, input parameters of network model were chosen as the height of convector, thickness of convector sheet, the trapezoidal height of convector, convector base length, opposing convector distance, tip width of convector, convector vertical location and distance between convectors. For the proposed neural network model, the mean squared errors obtained for the heat transfer and panel radiator weight are -1.25E-04 and -7.54E-05 respectively. In addition, an R-value of 0.99999 has been obtained, and the average deviation value has been calculated as 0.001%. The obtained results show that, depending on the design parameters, the proposed artificial neural network model can predict the rate of heat transfer and weight of the panel radiator with high accuracy. This investigation is supposed to fill a significant gap since it is the pioneer one in open sources on machine learning modeling of panel radiators. Thus, it can possibly make a crucial contribution to the related manufacturing industry.
Artificial Neural Networking (ANN) Model for Convective Heat Transfer in Thermally Magnetized Multiple Flow Regimes with Temperature Stratification Effects
(Mdpi, 2022) Rehman, Khalil Ur; Colak, Andac Batur; Shatanawi, Wasfi
The convective heat transfer in non-Newtonian fluid flow in the presence of temperature stratification, heat generation, and heat absorption effects is debated by using artificial neural networking. The heat transfer rate is examined for the four different thermal flow regimes namely (I) thermal flow field towards a flat surface along with thermal radiations, (II) thermal flow field towards a flat surface without thermal radiations, (III) thermal flow field over a cylindrical surface with thermal radiations, and (IV) thermal flow field over a cylindrical surface without thermal radiations. For each regime, a Nusselt number is carried out to construct an artificial neural networking model. The model prediction performance is reported by using varied neuron numbers and input parameters, and the results are assessed. The ANN model is designed by using the Bayesian regularization training procedure, and a high-performing MLP network model is used. The data used in the creation of the MLP network was 80 percent for model training and 20 percent for testing. The graph shows the degree of agreement between the ANN model projected values and the goal values. We discovered that an artificial neural network model can provide high-efficiency forecasts for heat transfer rates having engineering standpoints. For both flat and cylindrical surfaces, the heat transfer normal to the surface reflects inciting nature towards the Prandtl number and heat absorption parameter, while the opposite is the case for the temperature stratification parameter and heat generation parameter. It is important to note that the magnitude of heat transfer is significantly larger for Flow Regime-IV in comparison with Flow Regimes-I, -II, and -III.
Artificial Neural Networking (ANN) Model for Drag Coefficient Optimization for Various Obstacles
(Mdpi, 2022) Rehman, Khalil Ur; Colak, Andac Batur; Shatanawi, Wasfi
For various obstacles in the path of a flowing liquid stream, an artificial neural networking (ANN) model is constructed to study the hydrodynamic force depending on the object. The multilayer perceptron (MLP), back propagation (BP), and feed-forward (FF) network models were employed to create the ANN model, which has a high prediction accuracy and a strong structure. To be more specific, circular-, octagon-, hexagon-, square-, and triangular-shaped cylinders are installed in a rectangular channel. The fluid is flowing from the left wall of the channel by following two velocity profiles explicitly linear velocity and parabolic velocity. The no-slip condition is maintained on the channel upper and bottom walls. The Neumann condition is applied to the outlet. The entire physical design is mathematically regulated using flow equations. The result is presented using the finite element approach, with the LBB-stable finite element pair and a hybrid meshing scheme. The drag coefficient values are calculated by doing line integration around installed obstructions for both linear and parabolic profiles. The values of the drag coefficient are predicted with high accuracy by developing an ANN model toward various obstacles.
Carbon dot unravels accumulation of triterpenoid in Evolvulus alsinoides hairy roots culture by stimulating growth, redox reactions and ANN machine learning model prediction of metabolic stress response
(Elsevier France-Editions Scientifiques Medicales Elsevier, 2024) Awere, Collince Omondi; Sneha, Anbalagan; Rakkammal, Kasinathan; Muthui, Martin Mwaura; Kumari, Anitha R.; Govindan, Suresh; Colak, Andac Batur
Evolvulus alsinoides, a therapeutically valuable shrub can provide consistent supply of secondary metabolites (SM) with pharmaceutical significance. Nonetheless, because of its short life cycle, fresh plant material for research and medicinal diagnostics is severely scarce throughout the year. The effects of exogenous carbon quantum dot (CD) application on metabolic profiles, machine learning (ML) prediction of metabolic stress response, and SM yields in hairy root cultures of E. alsinoides were investigated and quantified. The range of the particle size distribution of the CDs was between 3 and 7 nm. The CDs EPR signal and spin trapping experiments demonstrated the formation of O2(-center dot)spin-adducts at (g = 2.0023). Carbon dot treatment increased the levels of hydrogen peroxide and malondialdehyde concentrations as well as increased antioxidant enzyme activity. CD treatments (6 mu g mL-1) significantly enhanced the accumulation of squalene and stigmasterol (7 and 5-fold respectively). The multilayer perceptron (MLP) algorithm demonstrated remarkable prediction accuracy (MSE value = 1.99E-03 and R2 = 0.99939) in both the training and testing sets for modelling. Based on the prediction, the maximum oxidative stress index and enzymatic activities were highest in the medium supplemented with 10 mu g mL-1 CDs. The outcome of this study indicated that, for the first time, using CD could serve as a novel elicitor for the production of valuable SM. MLP may also be used as a forward-thinking tool to optimize and predict SM with high pharmaceutical significance. This study would be a touchstone for understanding the use of ML and luminescent nanomaterials in the production and commercialization of important SM.
Comparative investigation of the usability of different machine learning algorithms in the analysis of battery thermal performances of electric vehicles
(Wiley-Hindawi, 2022) Colak, Andac Batur
The cooling and thermal management of battery packs, which are the most important components used in electric vehicles (EV), are of critical importance for the efficiency and performance of EV. This study aims to analyze the usability of machine learning algorithms in determining the thermal parameters of the battery thermal management system (BTMS) used in EV and to determine the machine learning algorithm with the highest prediction performance. The prediction performance of three different artificial neural networks developed by using Levenberg-Marquardt, Bayesian Regularization (BR), and Scaled Conjugate Gradient (SCG) machine learning algorithms have been extensively and comparatively investigated. As input parameters of the models, discharge rate, flow rate, and inlet temperature values were defined and the average temperature of the battery surface and maximum temperature difference on the surface values were estimated. The coefficient of determination values for the Levenberg-Marquardt, BR, and SCG algorithms was calculated as 0.99848, 0.98751, and 0.97592, respectively. The results showed that the machine learning algorithms can determine the thermal parameters of the BTMS of EV with high accuracy. However, it has been observed that the highest prediction accuracy belongs to the Levenberg-Marquardt algorithm.

Yazar "Colak, Andac Batur" seçeneğine göre listele

Sayfa Başına Sonuç

Sıralama seçenekleri