Layer Artificial Neural Network Model Research Articles

The Effect of Varying Artificial Neural Network and Adaptive Neuro-Fuzzy Inference System Parameters on Wind Energy Prediction: A Comparative Study

Owing to the development of technology, the majority of nations throughout the world now rely on fossil fuels and nuclear power plants to meet their energy needs. However, as academic research on this subject has shown, it has become clear that alternative energy uses are necessary due to the gradual depletion of these fuels and their significant negative effects on the environment. In order to ensure energy diversity and end the energy shortage, the development of renewable energy sources is crucial. The prediction of wind power is crucial for effectively utilizing the potential of wind energy. In this study, an adaptive neuro-fuzzy inference system (ANFIS) and an artificial neural network (ANN) have been developed for the prediction of wind power. In this study, data sets were created by taking the daily average wind speeds of the selected wind turbine, the daily average power values it produces, and the daily average wind speed values in the Velimese region. By creating single-hidden layer and multi-hidden layer ANN models, the network was trained multiple times with different activation functions and different numbers of neurons, and wind power prediction was performed. In the ANFIS model, the number of membership functions is kept constant, and wind power prediction is performed using different membership functions. With these ANFIS and ANN models developed with different parameter combinations, it is aimed to determine the most efficient model by performing daily average wind power prediction. Parameter combinations were tested to determine the appropriate models, and as a result, the ANN and ANFIS models were compared with each other.

Synthesis and analysis of TiO2 nanotubes by electrochemical anodization and machine learning method for hydrogen sensors

The conductometric hydrogen gas sensors were used to explore TiO2 nanotubes in this study. TiO2 nanotubes are synthesized by anodization of the titanium foils using a neutral 0.5% and 1% (wt) NH4F in glycerol solution depending on anodization time and anodization voltage at the temperature of 20 °C. The amorphous, rutile and anatase phases of TiO2 are observed for as-prepared TiO2 nanotubes, annealed at 700 and 300 °C, respectively. The diameters of the nanotubes grow as the anodization time and voltage increase, according to scanning electron microscopy (SEM) images. The inner diameter of nanotubes is changed between ~70 nm to ~225 nm. Hydrogen sensing properties of Ti/TiO2 nanotubes/Pd device has been tested at room temperature under concertation range from 0.5% to 10% depending on the crystalline phase. The highest sensor response is observed for anatase crystalline TiO2 nanotubes. Typical Schottky-type behavior is observed from the I-V measurement. All the fabricated nanotube diameters are also simulated by using Support Vector Machine and Artificial Neural Network models. And also, some of the nanotube diameters which are not obtained experimentally (anodization voltage of 70 V) are estimated using the Support Vector Machine and Artificial Neural Network models. In addition, an analytical model is also proposed using Jacobi numeric analysis method alternative to the simulation model for the nanotube diameter. Finally, the analytical, simulation, and experimental results are compared, and the best result is obtained using the 1 Hidden Layer Artificial Neural Network model.

Response Surface Methodology and Artificial Neural Network-Based Models for Predicting Performance of Wire Electrical Discharge Machining of Inconel 718 Alloy

This paper deals with the development and comparison of prediction models established using response surface methodology (RSM) and artificial neural network (ANN) for a wire electrical discharge machining (WEDM) process. The WEDM experiments were designed using central composite design (CCD) for machining of Inconel 718 superalloy. During experimentation, the pulse-on-time (TON), pulse-off-time (TOFF), servo-voltage (SV), peak current (IP), and wire tension (WT) were chosen as control factors, whereas, the kerf width (Kf), surface roughness (Ra), and materials removal rate (MRR) were selected as performance attributes. The analysis of variance tests was performed to identify the control factors that significantly affect the performance attributes. The double hidden layer ANN model was developed using a back-propagation ANN algorithm, trained by the experimental results. The prediction accuracy of the established ANN model was found to be superior to the RSM model. Finally, the Non-Dominated Sorting Genetic Algorithm-II (NSGA- II) was implemented to determine the optimum WEDM conditions from multiple objectives.

Application of artificial neural network for comparison and modeling of the ultrasonic and stirrer assisted removal of anionic dye using activated carbon supported with nanostructure material

In this study, a green approach has been described for the synthesis of copper sulfide nanoparticles loaded on activated carbon (CuS‐NP‐AC) and usability of it for the removal of sunset yellow (SY) dye by ultrasound‐assisted and stirrer has been compared. In addition, the artificial neural network (ANN) model has been employed for a forecasting removal percentage of SY dye using the results obtained. This material was characterized using scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The impact of variables, including initial dye concentration (mg/L), pH, adsorbent dosage (g), sonication time (min) and temperature (°C) on SY removal was studied. Fitting the experimental equilibrium data of different isotherm models such as Langmuir, Freundlich, Temkin and Dubinin–Radushkevich models display the suitability and applicability of the Langmuir model. Analysis of experimental adsorption data of different kinetic models including pseudo‐first and second order, Elovich and intraparticle diffusion models indicate the applicability of the second‐order equation model. The adsorbent (0.005 g) is applicable for successful removal of SY dye (> 98%) in short time (9 min) under ultrasound condition. A three layer ANN models with 8 and 6 neurons at hidden layer was selected as optimal models using stirrer and ultrasonic, respectively. These models displayed a good agreement between forecasted data and experimental data with the determination coefficient (R2) of 0.9948 and 0.9907 and mean squared error (MSE) of 0.0001 and 0.0002 for training set using stirrer and ultrasonic, respectively.

Design of a feature-tuned ANN model based on bulk rock-derived mineral spectra for endmember classification of a hyperspectral image from an iron ore deposit

In this article, a new feature-tuned artificial neural network (ANN) model has been developed for endmember classification of a hyperspectral image. This model is developed on the basis of using only the essential absorption bands of mineral spectra as opposed to using all the spectral bands of the hyperspectral image. This approach has the added advantages of reducing the dimensionality of input features to the ANN as well as inhibiting the influences of noisy bands for classification of endmembers. The proposed ANN model is trained using input features extracted from laboratory spectra of in situ bulk ore materials collected from an existing iron ore deposit. The input features are basically the constituent absorption bands of mineral spectra where each absorption band is mathematically characterized by the centre, width, and strength parameters of a Gaussian curve. For extracting absorption bands from a mineral spectrum, a modified Gaussian model has been used. The application of this model also necessitates the design of a special template for the input layer ANN model. After training the model, its generalization property is assessed through a testing data set. The model has achieved nearly 97% of classification accuracy in a training set, and 71% of accuracy in a testing set. The trained model is then applied on Hyperion imagery collected over an iron ore deposit. All the endmember spectra of this deposit are classified into either vegetation or any of the ores or rock present in the deposit. None of the endmembers is classified into non-iron ore minerals.

The photocatalytic degradation of carbamazepine and prediction by artificial neural networks

The three layer artificial neural network model was applied to predict the degradation efficiency for carbamazepine in photocatalytic oxidation under UV radiation. Titania–zirconia was employed as a catalyst for the photooxidation. The catalyst was prepared using titanium isopropoxide and zirconium oxychloride by sol–gel method and characterized by transmission electron microscopy and BET analysis. Different process parameters such as, initial concentration of carbamazepine, pH of the solution, catalyst concentration and time of UV irradiation were employed as the input to the artificial neural network model and the output of the network was degradation efficiency of carbamazepine. The multilayer feed-forward networks with the Levenberg–Marquardt (trainlm) backpropagation training algorithm was used for the network training. The smallest mean square error was obtained for three-layer network with ‘logsig’ transfer function and five neurons in the hidden layer gave optimal results. A comparison between the predicted values and selective experimental data of degradation efficiency showed a high correlation coefficient (R2) of 0.997.

Simulation and Optimization of Artificial Neural Network Modeling for Prediction of Sorption Efficiency of Nanocellulose Fibers for Removal of Cd (II) Ions from Aqueous System

Simulation and optimization of an Artificial Neural Network (ANN) for modeling biosorption studies of cadmium removal using nanocellulose fibers (NCFs) was carried out. Experimental studies led to the standardization of the optimum conditions for the removal of cadmium ions i.e. biomass dosage (0.5 g), test volume (200 ml), metal concentration (25 mg/l), pH (6.5) and contact time (40 min). A Single layer ANN model was developed to simulate the process and to predict the sorption efficiency of Cd (II) ions using NCFs. Different NN architectures were tested by varying network topology, resulting in excellent agreement between experiment outputs and ANN outputs. The findings indicated that ANN provided reasonable predictive performance for training, cross validation and testing data sets (R2 = 0.998, 0.995, 0.992). A sensitivity analysis was carried out to assess the influence of different independent parameters on the biosorption efficiency, and pH > biomass dosage > metal concentration > contact time > test volume were found to be the most significant factors. Simulations based on the developed ANN model can estimate the behavior of the biosorption phenomenon process under different experimental conditions.doi:10.14456/WJST.2014.4

Time–delay single layer artificial neural network models for estimating shelf life of burfi

Time-delay single layer artificial neural network models were developed for estimating the shelf life of burfi stored at 30oC. Input variables for the models were moisture, titratable acidity, free fatty acids, tyrosine, and peroxide value, while the overall acceptability score was taken as output variable. Mean square error, root mean square error, coefficient of determination and Nash - sutcliffo coefficient were applied in order to compare the prediction ability of the Time-delay single layer models. The combination of 5-1-1 showed a very high correlation between the training and validation data, establishing that the developed models are effective in predicting the shelf life of burfi.

Comparison of mouse embryo deformation modeling under needle injection using analytical Jacobian, nonlinear least square and artificial neural network techniques

Analytical Jacobian, nonlinear least square and three layer artificial neural network models are employed to predict deformation of mouse embryos under needle injection, based on experimental data captured from literature. The Maximum Absolute Error (MAE), coefficient of determination (R2), Relative Error of Prediction (REP), Root Mean Square Error of Prediction (RMSEP), Nash–Sutcliffe coefficient of efficiency (Ef) and accuracy factor (Af) are used as the basis for comparison of these three models. Analytical Jacobian, nonlinear least square and ANN models have yielded the correlation coefficient of 0.9985, 0.9964 and 0.9998, respectively. The REP between the models predicted values and experimental observations are 2.8228(%), 4.7647(%) and 0.4698(%) for the analytical Jacobian, nonlinear least square and ANN methods, respectively. Results showed that ANN performed relatively better than the analytical Jacobian and nonlinear least square methods. Findings indicate that the ANN technique can predict mouse embryo dimple depth by needle injection with considerable accuracy and the least error.

Artificial Neural Network (ANN) design for Hg–Se interactions and their effect on reduction of Hg uptake by radish plant

The tendency of selenium to interact with heavy metals in presence of naturally occurring species has been exploited for the development of green bioremediation of toxic metals from soil using Artificial Neural Network (ANN) modeling. The cross validation of the data for the reduction in uptake of Hg(II) ions in the plant R. sativus grown in soil and sand culture in presence of selenium has been used for ANN modeling. ANN model based on the combination of back propagation and principal component analysis was able to predict the reduction in Hg uptake with a sigmoid axon transfer function. The data of fifty laboratory experimental sets were used for structuring single layer ANN model. Series of experiments resulted into the performance evaluation based on considering 20% data for testing and 20% data for cross validation at 1,500 Epoch with 0.70 momentums The Levenberg–Marquardt algorithm (LMA) was found as the best of BP algorithms with a minimum mean squared error at the eighth place of the decimal for training (MSE) and cross validation.

Single hidden layer artificial neural network models versus multiple linear regression model in forecasting the time series of total ozone

Present paper endeavors to develop predictive artificial neural network model for forecasting the mean monthly total ozone concentration over Arosa, Switzerland. Single hidden layer neural network models with variable number of nodes have been developed and their performances have been evaluated using the method of least squares and error estimation. Their performances have been compared with multiple linear regression model. Ultimately, single-hidden-layer model with 8 hidden nodes have been identified as the best predictive model.

Comparative study of artificial neural network and statistical models for predicting student grade point averages

This paper compares linear regression; stepwise polynomial regression; and fully-connected, single middle layer artificial neural network models with an index used by an admissions committee for predicting student GPAs in professional school. It also provides methods for implementing, interpreting, and evaluating artificial neural networks, including an optimization of model structure for simple neural networks. While the neural network identifies additional model structure over the regression models, none of the empirical methods was statistically significantly better than the practitioners' index.