جستجوی مقالات مرتبط با کلیدواژه « الگوریتم ژنتیک » در نشریات گروه « مهندسی شیمی، نفت و پلیمر »

Saturation/flow functions, i.e., water and oil relative permeability and capillary pressure are the most critical input parameters that play a key role in porous media simulation studies and forecasting the fluids recovery. Various methods including laboratory tests (steady or unsteady core flooding), and mathematical methods can be utilized to measure or estimate the flow function curves. In this research, history matching technique was performed using optimization genetic algorithms to adjust the parameters of the proposed model based on laboratory water flooding data (pressure drop and cumulative oil production) and the results of core flooding simulation. To do so, multi-objective genetic algorithms were employed to find optimum solutions for piecewise model and then the results are compared with several empirical models (e.g. Corey and Brooks-Corey models). History matching results showed that single objective optimization, i.e., considering cumulative oil production as an objective function provides a good fit for Corey’s model (R-squared= 99.56%) and single and multi-objective optimization lead to the best fits, respectively, with the accuracies of 99.57 and 99.16% for piecewise model during the imbibition process. Moreover, the drainage history matching results showed that the piecewise model exhibits the best performance employing the single objective and multi-objective optimization algorithms with the accuracies of 99.32 and 99.85%, respectively. Overall, the piecewise model can be utilized to address the history matching problem using multi-objective optimization and estimate the flow functions curves. Hence, it is proposed as an efficient method for estimating fluid relative permeability and capillary pressure functions based on experimental and simulation data.

In the customer's interaction with the company, environmental information is received by the five senses. The compatibility or lack of compatibility of the environment's characteristics with the customer's sensory tastes can cause the customer's satisfaction or dissatisfaction. This research has two goals. The effect of mixed sensitive marketing on customer loyalty of Iranian clothing brands and optimization of the effect of mixed sensitive marketing on customer loyalty. The research method is descriptive-survey and practical in terms of purpose. The statistical population is unlimited and includes the buyers and consumers of clothes of 18 Iranian brands in the age group of children, teenagers and adults. The sample size of 384 was determined and selected using non-probability sampling method. In order to collect the data, the mixed sensitive marketing questionnaires of Kian and Loui 2009 and the customer loyalty questionnaire of vaida,kanopaite 2015 were used. In order to analyze the data, structural equations and LISREL software were used. The meta-heuristic algorithm part of the genetic algorithm was used for its optimal modeling. In this way, the rules related to the effect of the mixed dimensions of sensory marketing on customer loyalty were first identified by the decision tree and then these rules were optimized with the help of genetic algorithm with MATLAB software. The results of the research in the first part showed the approval of all sensory marketing components except the interaction component. In the second part, using the decision tree with the help of the genetic optimization algorithm, customer loyalty can be predicted with 97% accuracy and accuracy using 27 rules.

Assessment of petroleum generation potential of the source rock as a function of total organic carbon content and kerogen type is of great importance in oil and gas exploration studies. The main aim of this research is to compare the performance of artificial neural networks trained by back propagation algorithm (ANN-BP) and metaheuristic methods including genetic algorithm (ANN-GA) and particle swarm optimization (ANN-PSO) for prediction of total organic carbon (TOC) content and remaining petroleum potential (S2) from the wireline data. For this purpose, Pabdeh Formation (Paleocene-Oligocene) in Mansuri oilfield is studied. Based on the results of linear regression on the test data, ANN-PSO method provides more accurate predictions of Rock-Eval derived TOC and S2 parameters with correlation coefficient (R2) values of 0.8548 and 0.9089, respectively. In addition, hydrogen index (HI) is appropriately predicted based on the relationship between TOC and S2 values obtained from the ANN-PSO method with R2 value of 0.6882, from which different types of kerogen can be distinguished with classification accuracy of 74 percent. Geochemical zonation of Pabdeh Formation based on organic richness and kerogen type reveals three distinctive parts, among which the middle part (Brown Shale Unit, BSU) demonstrates the greater petroleum generation potential with having the significant values of total organic carbon and hydrogen index. Therefore, the BSU can play an important role in hydrocarbon charging of the oilfield traps if it attains proper level of thermal maturity. Accordingly, precise determination of petroleum generation characteristics of Pabdeh Formation using the ANN-PSO model proposed in this study will lead to a reduction in uncertainty associated with petroleum system modeling, and therefore will considerably increase the exploration efficiency in the Mansuri oilfield.

After recognizing the toxic and carcinogenic effects of Lead organic compounds, production of compounds such as Methyl tertiary butyl ether as an additive to ordinary hydro carbonate gases was proposed. These materials are able to enter into water, soil and air due to relative steam pressure and partial dissolution in water and cause the pollution. As a result, development of a new process for producing gas with high octane from complex compounds of light petroleum distillates was initiated. This method is based on separating C5-C8 linear and branched alkanes according to their absorption properties, chain length and the number of branches. In this study, the hybrid neural network model based on experimental data in the database has been used as an alternative model for predicting the separation rate of linear and branched paraffin through absorption process. Absorption temperature, absorption time, hydrocarbons' octane number, and hydrocarbon density are considered as four input parameters, and the ratio of linear paraffin concentration to total (C/C0) as the output parameter of neural network. The neural network model was successfully generalized by experimental database and then was investigated with the help of test data. The results of modeling for the test data indicated the success of neural network model in predicting the rate of linear paraffin separation from non-linear ones. Therefore, the developed neural network model can be used for determining the C/C0 with confidence in absorption process. According the obtained results for test data, the minimum mean squared error is 0/0518, which is a satisfactory measure. The model and experimental data were compared and regression coefficient 0.990 shows good matching between modeling results and experimental results.

In this study, the process of single stage mixed natural gas liquefaction is reviewed, then process energy optimization and economic analysis are followed. The effect of specific energy consumption optimization on the costs of cooling equipment was investigated. The basis of this study is PRICO process and due to some limitations and concerns for Iran, it is tried to develop a new modified process so that the available process restrictions would expel and finally the recommended process optimized by genetic algorithm. Mole flow of refrigerant, low pressure and high pressure of refrigerant in the cycle, are the most important operational parameters that were selected by sensitivity analysis. The total energy consumption of compressors reached 1413.8 kW and decreased by 11 percent. Results showed optimization of the specific energy consumption increases the cost of the heat exchanger, due to the reduction of heat exchanger temperature driving force and need to higher heat transfer area. However, the operating cost of compressors can be reduced.

Addition of nanoparticles to a polymeric matrix leads to enhance the performance of membrane gas separation. In this study, the aim is to find the optimum operative point of polymeric membrane modified by adding nanoparticles in gas separation. The assessed factors are type of nanoparticle, percentage of added nanoparticle, and cross membrane pressure.  Nanoparticles of AL2O3, ZnO, and TiO2 were used. Further, the ranges of nanoparticle concentration and operative cross membrane pressure were 2.5 to 15% and 2 to 25 bar respectively. To optimize a process, developing a robust model is necessary. Therefore, first, a powerful model based on artificial neural network was developed, which it was able to predict the values of permeability of oxygen, nitrogen, methane, and carbon dioxide. Neural network models were developed that had R2 greater than 0.9. Next, the optimum operative conditions for assessed gases were found using methodology based on genetic algorithm and considering four strategies. The results of optimization show that the maximum values of permeability for oxygen, nitrogen, methane, and carbon dioxide are 334.7, 779.9, 902.7, and 270.4 respectively.

Initially a model of neural network analysis achieved and was using by genetic algorithm optimization. Comparing the results of Petro-SIM’s optimization software, and genetic algorithm showed that the temperature in both are equal and pressure from genetic algorithms and Petro-SIM’s software, respectively 75.6 psig and 79.5 psig that are slightly lower than the amount of operational. In contrast the help of genetic algorithm to optimize air flow rate is 181 kg/hr and the optimum value obtained from the Software is 169.8 kg/hr. In addition, optimum weight fraction of soda in both techniques is almost equal to the values in the unit. Using the results of genetic algorithm and Entering this condition to Petro-SIM software and simulation it became clear that optimize the optimum conditions, the value of mercaptan concentration in the product will be less than 1ppm.

In recent years, the capabilities of rotating packed beds in dust removal from gases has attracted the attention in different industries. On the other hand, due to intricate complexities of such beds, employment of computational fluid dynamic techniques for design and modeling purposes would be difficult. Therefore, in this research the performance of artificial intelligence techniques in design of theses equipment was investigated. Accordingly, in the first step and by using 561 experimental data points, an artificial neural network system was developed. In the constructed model, dust removal efficiency was considered as the output and the working and design conditions of the bed were fed as the inputs. Performance of the developed system was analyzed both quantitatively and qualitatively and achieving value of 0.99 for correlation coefficient demonstrates its suitable accuracy in efficiency determination. Finally, employing the developed neural network system combined with genetic algorithms, a novel method for optimum design of beds with arbitrary characteristics was proposed. Performed analyses showed that the constructed system while being simple, offers flexibility and precision in design of rotating packed beds. Also, its low coast of development makes it an attractive idea to be employed in industry scale.

Recently, optimization methods have been extensively applied in phase equilibrium calculations. Among these methods, Genetic Algorithm (GA) can be used to calculate the interaction parameters of activity coefficient models in equilibrium systems. In this study, based on the genetic algorithm, the interaction parameters of 5 activity coefficient models (2-suffix Margules, 3-suffix Margules, Wilson, NRTL and UNIQUAC) have been calculated for 20 ternary extraction systems (water + carboxylic acids + organic solvents) including 126tie-lines. The values of binary interaction parameters of these models along with the root mean square deviations (RMSD) are reported. The mean values of RMSD of the systems in the order of the mentioned models have been calculated 0.0298, 0.0067, 0.0114, 0.0025 and 0.0052, respectively. The results show that all models except 2-suffix Margules model have relatively good accuracy. By comparing the RMSD values in literature and the values determined by GA for NRTL and UNIQUAC models, the RMSD values of the models improved from 0.0124 and 0.0181 to 0.0025 and 0.0052, respectively

One of the non-destructive tests developed for determining fruit maturity stage is the analysis of imapact fruit response of fruit . Therefore, in this study, in order to evaluate the quality and maturity stages of pomegranate fruit, a system for recording acoustic responses was developted. For this purpose, the acoustic response of 150 pomegranates in three maturity stages (immature, ripe, and over-ripe) was recorded. To evaluate the destructive properties, the total soluble solids, acidity, ascoric acid , anthocyanin, total phenol and pH were measured; Also, in order to record and extract the acoustic features (non-destructive test), positions of microphone (behind, next to and under the fruit), as well as the hitting angle (5, 10 and 15 degrees) were evaluated as variables. In order to classify the data, a decision tree classifier combined with the genetic algorithm was utilized. Based on the statistical analysis, it was determined that the best orientation of the microphone and the hitting angle were the placement of the microphone next to hitting ball and 15 degrees, respectivily. Based on the results of the classification, the highest, lowest and overall classification accuricy were 96.7% (in the ripe group), 73% (in the over-ripe group) and 89.2%, respectively.

Mixed refrigerant is widely used in the natural gas liquefaction process, and this makes it important to optimize the parameters affecting this process. In this research, a PRICO process with mixed refrigerant was optimized by using Non-dominated Sorting Genetic Algorithm (NSGAII). In this optimization, energy consumption and the surface of heat exchanger were optimized simultaneously to eliminate the adverse effect of the surface of heat exchanger increase due to the single-objective optimization of the energy consumption observed in the previous papers. Molar flow rates of mixed refrigerant compounds including methane, ethylene, propane, isopentane and nitrogen, outlet pressures of the first and second compressors and outlet pressure of the cooling cycle valve were considered as optimization variables. Also, the considered objective functions were performed on sensitive analysis towards optimization variables Moreover, the optimization solutions were presented as a set of optimal solutions called Pareto optimal fronts whose results showed the contradiction of the two objective functions well. Afterwards, multi-objective optimization results were compared with single-objective minimization results. Ultimately, comparison results showed that the surface of heat exchanger in multi-objective optimization was reduced by 54% in comparison with  single-objective optimization (energy optimization).

Black mulberry is a very important and rich source of polyphenols and anthocyanin’s that dry and maintain can be valuable. The use of infrared radiation as a new method in drying preserves the quality of the final product, which increases the speed of drying and reduces the cost of the process. The effect of 120, 180, 240 W of infrared power, 5, 10 and 15 cm of sample distance from infrared source and drying time on black mulberry drying were investigated. Modeling of black mulberry drying was used to predict the output of this study using genetic algorithm-artificial neural network with four inputs (infrared power, sample distance from infrared source and drying time) and one output (weight loss percentage). It was determined that by increasing the infrared power and reducing the distance between the samples and the infrared lamp, the drying rate of black mulberry was significantly increased by two-way infrared (P<0.05). The sigmoid activation function was selected as the activation function in the hidden layer and output layer due to the lower error value than other functions. According to the test method and the fault method, 15% of the data was used for training to achieve the best learning conditions for the relationships between inputs and outputs by the network. 15% of the data was used for the trained network test, and 60% of the remaining data was used to evaluate the network. According to the results, it can be concluded that using the sigmoid activation function and the network with 8 neurons in a hidden layer, we can well analyze the percentage of weight loss using the genetic algorithm-artificial neural network during the black mulberry drying process was predicted by two-way infra-red method (R2=0/999). Drying time was introduced as the most effective factor for controlling black mulberry weight loss using sensitivity analysis by optimal neural network.

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.

One of the most important sections in fulfilled study (FFS) and master development plan is history matching which plays an important role in production scenarios and future production plan of reservoir. It would be a challenge for reservoir engineering due to lots of parameters and uncertainties during reservoir study which need lots of simulation runs to reach good match for responses in conventional mechanism of history matching. However, for accelerating history matching part, new methods which called as assisted or automated history matching (AHM) have been established. In this paper, an approach for automated history matching (AHM) was applied in a real brown field with 14 wells with multiple responses located in south of Iran. The main important features of the proposed algorithm were defining a proxy model which was a response surface method in which 21 model parameters were incorporated based on cubic centered face method. Response surface method (RSM) has been employed to create the proxy model. The optimization algorithms utilized in this research were genetic algorithm (GA) and particle swarm optimization (PSO). Proxy model was successfully constructed using 256 samples leading into R2 of 0.9 on dataset with good results on GA.

In this research, the application of design of experiment and artificial neural network on conversion of H2 and CO were studied based on the experimental data. The experimental data has been collected from five independent variables based on central composite design such as temperature and pressure of reactor, H2/CO feed ratio, and partial pressure of H2 and CO in reactor. The operating conditions are: T = 320-340°C, P = 2-8 barg, H2/CO = 0.8-2.2, PCO = 0.3-2.7 barg, and PH2 = 0.3-2.5 barg. To generate the conversion models, two methods consist of response surface methodology and artificial neural network were used. The capability and sensitivity of both models were evaluated by some statistical parameters including mean square error and absolute average relative deviation. The result of both models were compared with experimental data and show the best results. To evaluate the maximum conversion of (H2 and CO), a hybrid ANN/GA was performed to solve the nonlinear both models. Finally, all quadratic equations and maximum of both models were performed, and the results were concluded. Also, this method can be used to produce the valuable selective production.

In general, the non-sharp distillation columns sequence optimization problem is formulated as a MINLP with a non-linear objective function. This study is an attempt to provide a two- phase approach which categorizes the variables into two classes in order to simplification of the problem. A set of variables are managed and optimized by Genetic Algorithm, the other set of the remaining variables can be solved as a LP problem for estimated values of the first set. The study formulates some mathematical relationships between the first set of variables from the non-linear constraints so that by optimizing some of them, others are obtained with the assistance of the relationships. The method has the ability to obtain the best structure from the overall superstructure directly in short time without using any starting points for ternary separation. Three published examples were used to validate the proposed method.

History matching is still one of the main challenging parts of reservoir study especially in giant brown oil fields with lots of wells. It would be a challenge in reservoir engineering that due to various parameters and uncertainties in study of reservoirs, many simulation runs are needed to reach a good match for responses in conventional mechanism of history matching. However, for accelerating history matching part, new methods, which are called as assisted or automated history matching (AHM), have been established. In this paper, the latest approach for automated history matching (AHM) has been applied in a real brown field containing 14 wells with multiple responses that is located in south of Iran. Least square support vector machine (LSSVM) has been applied to create proxy model based on cubic centered face method. The optimization algorithms, used in this research, consist of genetic algorithm (GA) and particle swarm optimization (PSO).  Introduction In the latest studies in geosciences and reservoir characterization, employing a proxy model that acts faster, instead of real reservoir model, has led to good results. One of the most important sections in fulfilled study (FFS) and master development plan is history matching, which plays an important role in production scenarios and future production plan of the under study reservoir. In this paper, one of the newest methods is used for making proxy model and then, the model for history matching is optimized.  Methodology and Approaches Least square support vector machine (LSSVM) has been employed to create proxy model based on cubic centered face (CCF) method. The optimization algorithms of genetic algorithm (GA) and particle swarm optimization (PSO) have been used in this research.  Results and Conclusions A new proxy model has been successfully constructed using 1086 samples leading into determination coefficient (R2)

Prediction of solubilities of industrial solid compounds in supercritical carbon dioxide (SC-CO2) have been a challenge since the last two decades. In this study, the Peng-RobinsonStryjek-Vera equation of state (PRSV EoS) and the SoaveRedlich-Kwong equation of state (SRK EoS) were applied to estimate the solubilities of 10 common solid compounds in supercritical carbon dioxide. Simulation results showed that the PRSV EoS had good agreement with experimental data. Both EoS needed binary interaction parameter for prediction of phase behavior of solid-supercritical. In addition, with use of genetic algorithm, parameter for solid compounds and SCCO2 were calculated. Due to the lack of comprehensive correlations for prediction of solubilities of CO2 solids, in this work a correlation to predict the solubility of solid in SCCO2 was proposed. The main advantage of this correlation is its simplicity and having only two inputs, namely temperature and pressure. Minimum value of the average absolute percent deviation (AAPD) was for PRSV EoS and carbon dioxidenaphthalene system (2.77%) and maximum value of this parameter was for SRK EoS and carbon dioxide- anthracene system (19.38%).

According to the benefits of oxidative desulfurization (ODS) and ultrasound assisted oxidative desulfurization (UAOD) processes and in order to industrialize these processes, it is necessary that the kinetics of these processes be studied and modeled, so as to be able to design and simulate them in semi industrial and industrial scale. Therefore, kinetics modeling of oxidative desulfurization reaction of 4,6-DMDBT over molybdenum catalyst supported on alumina has been investigated. For expression of reaction kinetics equations, Langmuir Hinshelwood and Eley Rideal mechanisms have been used. Genetic algorithm was used for optimization. Based on, Langmuir-Hinshelwood and Eley Rideal models results, Langmuir-Hinshelwood mechanism can describe the kinetic behavior of UAOD and ODS processes. Micro kinetics equations and rate determining step are identified according to the results of the models.