kamyar movagharnejad

This research aimed to estimate how sample thickness (3, 5, and 7 mm), air velocity (0.5, 1.25, and 2 ms^(-1)), and infrared lamp power (1000, 1500, and 2000 W) affected the drying characteristics of zucchini slices in a far-infrared dryer. The design of drying tests, which illustrates the relation between input and output variables, was carried out by response surface methodology. Then, the effect of these independent factors on the moisture content (drying time), water activity, total phenolic content, potassium content and total color change was determined. To predict the responses, the most accurate mathematical models were selected using analysis of variance. The results revealed the drying time declined by decreasing the sample thickness and air velocity and increasing the infrared lamp power. Total phenolic content and potassium content had increased multifold after the drying process. An enhancement in sample thickness and infrared lamp power had a positive effect on both of these contents. The minimum and maximum values of total color change were 5.54 and 13.96, respectively. The dried zucchini produced by far-infrared drying is a high-value product that has the potential to be used in food and pharmaceutical industries or in the treatment of diseases as a booster meal.

The accurate description of the phase equilibria of CO2 and n-alkane multicomponent mixtures over a wide range of temperature, pressure, and n-alkane molecular weight, requires the models that are both consistent and mathematically flexible for such highly non-ideal systems. In this study, a predictive correlation was proposed for the vapor-liquid equilibrium data (VLE) of CO2 and n-alkane ternary systems, based on the Peng-Robinson equation of state (PR EOS), coupled to cubic mixing rules (CMRs). The ternary interaction parameters (TIP) correlation is developed using binary VLE data and tested for CO2 + n-alkane+ n-alkane ternary systems. For this purpose, binary VLE data of CO2 + n-alkane and n-alkane + n-alkane systems for n-alkane from C3 to C24, covering a total of about 70 references, used to correlate TIP in the ranges of 0.5-31 MPa and 230-663 K. Two temperature-dependent TIP correlations, based on acentric factor ratio, have been tuned with more than 2000 data points of the CO2 + n-alkane and the n-alkane + n-alkane binary systems with AARD of 3.13% and 6.71%, respectively. The generalized predictive correlation was proposed based on the proper three-body interaction contributions and successfully tested for VLE data of the CO2 + n-alkane + n-alkane ternary systems.

Rice is one of the most important cereals and is the second-highest worldwide production after wheat and also is a good source of nutrients for humans. It plays an important role in the feeding of the many parts of the world including Iran. The harvested paddy rice has the high initial moisture content of nearly 25-28% (wet basis) that caused corruption. Therefore, in order to prevent corruption and safe storage, it must be dried to 10-13% moisture content. Drying is one of the oldest methods of preserving food and agricultural products that used to increase the food’s storage time. There are several methods for drying paddy rice that none of them are ideal and have several advantages and disadvantages that one of them that recently the use of it has been increased is microwave drying. Microwave drying uses electromagnetic radiations with the frequency range of 300 MHz to 300 GHz and the wavelengths of 1-0.01m. In microwave drying due to better energy concentration, moisture is removed more quickly so the drying time decreases. Due to the complex relationship between input and output variables in the drying process, the selection of the model that can estimate the drying behavior of the products is difficult. Hence, the use of intelligent modeling methods such as neural networks is the best choice.

In this research, in order to investigate the effect of microwave power on kinetics of rice drying, head rice yield and effective diffusivity coefficient of moisture, a continuous type of domestic microwave dryer ( DEM-281 QOT-PW) were used. This dryer has a microwave radiation chamber where the samples are put on it on the tray that was placed on a digital balance. The experiments were performed at three microwave power levels designated as 270, 360 and 450 W. Also, Shirudi paddy rice was used as the raw material and the drying rice process from the initial moisture content of 21% to the final moisture content of 11% is examined. In this study, the neural network toolbox of MATLAB 2017R was used to model the kinetics of rice drying in the microwave dryers. RBF and MLP have 3 layers including input, hidden and output layers. The input layer has two neurons that show the number of input variables that were time and microwave power and the output layer has one neuron that shows the number of output variables that was MR in this study. 70% and 30% of the data was used for training and testing the network, respectively. To estimate the ANN performance, mean square error (MSE) and the coefficient of determination (R2) was used.

The maximum and minimum drying time was 42 and 20 minutes in 270 and 450 watts, respectively. Also, the maximum and minimum effective diffusivity coefficient of moisture were 4.17 * 10^-9 and 1.82* 10^-9 in 450 and 270 watts, respectively. RBF network with Guassian transfer function and high neurons number and MLP network with Levenberg-Marquardt ( LM) learning algorithm and tan-sigmoid (tansig) transfer function with low neurons number were able to model the kinetics of drying as well as. In general, the drying time and head rice yield decreased but the effective diffusivity coefficient of moisture increased by increasing the microwave power. Also drying at different microwave power did not affect rice color and quality.  The results of the modeling of rice drying by using two different neural networks including MLP and RBF demonstrated that the MLP network with Levenberg-Mrrqurdt (LM) learning algorithm and tan-sigmoid (tansig) transfer function has the better performance than the RBF network with Gussian transfer function and the error and the correlation coefficient in MLP are less and higher than the RBF, respectively.

Bubble point pressure is an important parameter in equilibrium calculations of reservoir fluids and having other applications in reservoir engineering. In this work, an artificial neural network (ANN) and a least square support vector machine (LS-SVM) have been used to predict the bubble point pressure of reservoir fluids. Also, the accuracy of the models have been compared to two-equation state-based models, i.e. SRK-EOS and PR-EOS and four empirical equations, i.e. Whitson, Standing, Wilson and Ghafoori et al. Compared to the experimental data, the average relative deviations (ARD) of bubble pressure prediction for these equations were obtained to be 14%, 29%, 66%, 30%, 38%, and 11%, respectively. The best semi-empirical equation has an ARD of about 11% while, the ANN and LS-SVM models have an ARD of 8% and 4.68%, respectively. Thus, it can be concluded that generally, these soft computing models appear to be more accurate than the empirical and EOS based methods for prediction of bubble point pressure of reservoir fluids.

Drying process of thin layer hawthorn fruits was investigated in a tray dryer in 4 different temperatures (60,65,70, 75 C) and 4 different air velocities ( 5.0, 6.0, 7.0, 8.0 m/s). The hawthorn samples were dried from original moisture content of 44.37% to final moisture content of 1.63%. The experimental data were fitted to different empirical correlations and the Genetic Algorithm optimization method was used to obtain the best parameters. The results showed that the modified Page model was the best model to describe the thin layer drying of hawthorn fruits in a tray dryer. The parameters of the empirical correlations may be presented as a function of temperature and velocity of the drying air. So, the generalized empirical correlations may describe the drying behavior of thin layer hawthorn in various conditions, while the previous empirical correlations were just applied to a certain drying experiment.

Neural network is one of the new soft computing methods commonly used for prediction of the thermodynamic properties of pure fluids and mixtures. In this study, we have used this soft computing method to predict the coefficients of the Antoine vapor pressure equation. Three transfer functions of tan-sigmoid (tansig), log-sigmoid (logsig), and linear were used to evaluate the performance of different transfer functions to redict the coefficients of the Antoine vapor pressure equation. The critical pressure, critical temperature, critical volume, molecular weight, and acentric factor were considered as the input variables and the Antoine equation coefficients showed by the symbols A, B, and C were considered as the output variables. The results of this study indicated that the linear transfer function had a better performance than other transfer functions and the topology of 5-6-3 with Levenberg–Marquardt learning algorithm and linear transfer function had the best performance for prediction of these coefficients.

Harvested rice has a high moisture content and it should be reduced to avoid corruption. In this study, the hot air drying was used to evaluate the kinetics of drying. This research was carried out under three variables including hot air velocity ( 0.65 , 0.8 m/s), hot air temperature ( 50 , 65 0C), and final moisture content (11, 13%). Modeling of rice drying was done by a multilayer perceptron artificial neural network. In order to evaluate the performance of training algorithms and transfer functions in predicting the drying behavior of paddy rice, three algorithms including Levenberg Marquardt, Resilient Bach Propagation and Scale Conjugate Gradient and two transfer functions including logsig and tansig were used. The results showed that the maximum drying time was approximately 8 hours at temperature of 500C, hot air velocity of 0.65 m/s and final moisture content of 11% and the minimum drying time was 1.21 hours at temperature of 650C, hot air velocity of 8 m/s and final moisture content of 13%. In general, the drying time decreased with increasing the temperature, hot air velocity and final moisture content. Also, the results of modeling showed that the levenberg Marquardt training algorithm had the best performance compared to the other algorithms. In general the topology of 3-11-1 with levenberg Marquardt training algorithm and logsig transfer function had the lowest mean square error and the highest correlation coefficient.

In this study, the refractive index of pure alcohols and binary alcohol mixtures were investigated experimentally and theoretically. In the experimental approach, a refractometer was used to measure the refractive index of the samples. Aa artificial neural network in form of a multilayer perceptron was also used to model and predict the measured refractive data. The input parameters of the network included temperature, molecular weight, the group contributions of CH3, CH2 and OH for the pure materials and for the binary mixtures the additional parameters of mole fractions, molecular weight and group contributions of both components have to be considered. The refractive index of the pure or binary alcohol mixture consist the only output parameter of the network. 70% of the experimental data were considered for train, 15% for test and 15% for the validation of the neural network. The optimum neural architecture for the pure compounds consisted of 10 neurons in the hidden layer with 0.08457 mean absolute relative error and the optimum network for binary mixtures consisted of 12 neurons in the hidden layer with 0.07121 absolute relative error. Comparison of the results showed a good agreement between the experimental data and the neural network outputs and the high accuracy of the model.

In this work, a Genetic Algorithm boosted Least Square Support Vector Machine model by a set of linear equations instead of a quadratic program, which is improved version of Support Vector Machine model, was used for estimation of 98 pure compounds second virial coefficient. Compounds were classified to the different groups. Finest parameters were obtained by Genetic Algorithm method for training data. The accuracy of the Genetic Algorithm boosted Least Square Support Vector Machine was compared with four empirical equations that are well-known and are claimed can predict all compounds second virial coefficients (Pitzer, Tesonopolos, Gasanov RK and Long Meng). Results showed that in all classes of compounds, the Genetic Algorithm boosted Least Square Support Vector Machine method was more accurate than these empirical correlations. The Average Relative Deviation percentage of overall data set was 2.53 for the Genetic Algorithm boosted Least Square Support Vector Machine model while the best Average Relative Deviation percentage for empirical models (Tesonopolos) was 15.38. When the molecules become more complex, the difference in accuracy becomes sharper for empirical models where the proposed Genetic Algorithm boosted Least Square Support Vector Machine model have predicted good results for classes of compounds that empirical correlations usually fail to give good estimates.

A total of 1099 data points consisting of alcohol-alcohol, alcohol-alkane, alkane-alkane, alcohol-amine and acid-acid binary solutions were collected from scientific literature to develop an appropriate artificial neural network (ANN) model. Temperature, molecular weight of the pure components, mole fraction of one component and the structural groups of the components were used as input parameters of the network while the refractive index was selected as its output. The ANN was optimized once by genetic algorithm (GA) and once again by particle swarm optimization algorithm (PSO) in order to predict the refractive index of binary solutions. The optimal topology of the ANN-GA consisted of 13 neurons in the hidden layer and the optimal topology of the ANN-PSO consisted of 16 neurons in the hidden layer. The results revealed that the ANN optimized by PSO had a better accuracy (MSE=0.003441 for test data) compared to the ANN optimized with GA (MSE=0.005117 for test data).

The water quality of the Karaj River was studied through collecting 2137 experimental data set gained by 20 sampling stations. The data included different parameters such as T (temperature), pH, NTU (turbidity), hardness, TDS (total dissolved solids), EC (electrical conductivity) and basic anion, cation concentrations. In this study a multi-layer perceptron artificial neural network model was designed to predict the calcium, sodium, chloride and sulfate ion concentrations of the Karaj River. 1495 data set were used for training, 321 data set were used for test and 321 data set were used for validation. The optimum model holds sigmoid tangent transfer function in the middle layer and three different forms of the training function. The root mean square error (RMSE), mean relative error (MRE) and regression coefficient (R) between experimental data and model’s outputs were measured for training, validation and testing data sets. The results indicate that the ANN model was successfully applied for prediction of calcium ion concentration.

Numerous empirical correlations exist for the estimation of crude oil viscosities. Most of these correlations are not based on the experimental and field data from Iranian geological zone. In this study several well-known empirical correlations including Beal, Beggs, Glasso, Labedi, Schmidt, Alikhan and Naseri were optimized and refitted with the Iranian oil field data. The results showed that the Beal and the Labedi methods were not suitable for estimation of the viscosity of the Iranian crudes, while the Beggs, Glasso and Schmidt methods gave reasonable results. The Naseri’s correlation and their present method proved to be the best classical methods investigated in this study. Two new intelligent methods to predict the viscosity of Iranian crudes have also been introduced. The study also showed that the neural network and SVM give much better results comparing to classical correlations. It is claimed that this study may provide more exact results for the prediction of Iranian oil viscosity.

In this research, the new microbial Bacillus licheniformisstrain ACO4 is used to produce a bioemulsifier. Based onlaboratorystudiesperformed, this bioemulsifier has a high emulsification capacity for the formation of the Iranian Nowruzheavy oil in water emulsion. This strain was able to reduce the viscosity of the emulsion from 10000 cP to 830 cP at pilot scale. This emulsion was stable for 72 h under the optimum conditions of temperature, water content and emulsifier concentration. The emulsion and heavy oil were passed through the pilot pipeline and the amount of sediment was observed and compared. The substantial stability of the oil in water emulsion allows the heavy oil to be transported practically over long distances or remain stationary for a considerable period of time prior to utilization.