genetic algorithm

Drying is one of the ways of storing of persimmon. In this study, to increasing shelf life of persimmon and producing high-quality products, infrared dryer was used and mass transfer kinetics, density, rehydration and color of samples were measured. The results showed that radiation lamp power and distance of lamp from sample had significant effect on the moisture loss kinetics and drying time (P<0.05). With increasing in radiation power, as well as reducing the distance of samples from the source of radiation, drying time decreased. The average density and rehydration for the dried samples in infrared were 639 kg /m3  and 270 %, respectively. The average calculated color changes (ΔE) for the power of 200, 300 and 400 w were 14.43, 10.09 and 20.04, respectively. The results of modeling by genetic algorithm-artificial neural network showed that artificial neural network combined with genetic algorithm provides better results and with combine them the speed of analysis and accuracy of modeling process increases. Using a network with 15 neurons in the hidden layer and using the hyperbolic tangent activation function and percentage data used to training/validation/testing equal 20/20/60 may be predicted drying kinetics of persimmon.

One of the oldest methods for the preservation of food is drying, which consists in removing water from the product in order to provide microbiological safety and the most popular drying method includes convection (Hassan-Beygi 2009; Salehi et al. 2014). One of the ways to shorten the drying time is to supply heat by infrared radiation (IR). Application of infrared heating to food drying is recently of special interest because of the progress in radiator construction. Their efficiency is between 80% and 90%, the emitted radiation is in narrow wavelength range and they are miniaturized (Pan et al. 2008; Salehi et al. 2014). Modeling of agricultural products drying is one of the best ways to controlling the drying time and conditions. Artificial neural networks (ANNs) are information processing networks constituting a set of highly interconnected neurons arranged in multiple layers that can be trained to fit one or more dependent variables to any degree of accuracy using a set of independent variables as inputs. ANNs were successfully applied to problems from various areas including business, medical and industrial fields. Once the ANN is trained using experimental data, it can be used in a purely predictive mode to calculate the dependent variable(s) for any values of input variables. Process modeling is an area where ANNs of various configurations and structures have been considered as alternative modeling techniques, particularly in cases where reliable mechanistic models cannot be obtained. Genetic algorithm (GA) optimization technique can be used to overcome this inherent limitation of ANN. GA are search techniques for an optimal value, mimicking the mechanism of biological evolution. They have a high ability to find an optimal value (global optimal value or at least near global one) of a complex objective function, without falling into local optima. The mathematical chromosomes could be operated upon by quasi genetic operations of selection, crossover, and mutation. These three parameters are repeated until desired convergence on optimal or near-optimal of the solutions is achieved. Mutation is a genetic operator that alters one or more gene values in a chromosome from its initial state. Mutation enhances the GA ability by intermittently injecting a random point in order to better search the entire parameter space, which allows the GA to possibly escape from local optima. This can result in entirely new gene values being added to the gene pool. With these new gene values, the GA may be able to arrive at a better solution than was previously possible (Ramzi et al. 2015; Salehi and Razavi 2016). Fresh apricots were obtained from market. Slices of apricot with 5 mm thickness were prepared with the aid of a steel cutter and were immediately placed into the dryer. The apricot slices were dried in an infrared dryer. In this study, for drying and increasing the shelf life of apricot, infrared radiation method was used. The effect of infrared lamp power at three levels 150, 250 and 375 watts, the distance of lamp from sample at three levels 5, 7.5 and 10 cm and time of 160 minute on drying of apricot were examined. Modeling of process was done with the genetic algorithm–artificial neural network (GA-ANN) method with 3 inputs (lamp power, distance and time) and 1 output (weight loss). During drying, radiation properties of the material are changing due to decreasing water content. As a consequence, its reflectivity increases and the absorptivity decrease. Generally, solid materials absorb infrared radiation in a thin surface layer (Pan et al 2008). The effects of infrared power and distance on the moisture content of apricot slices are shown in Figures 3 and 5. As expected, the moisture content was decreased by increasing the power because of the
increased temperature and heat transfer gradient between the air and samples. The weight loss of apricot samples were 36.34, 52.00 and 87.90 min at 150, 250 and 375 W, respectively (32 min and 10 cm). In conclusion, experimental results showed that the infrared power has a significant effect on the evolution of moisture content. Hence, drying of thin layers seems to be more efficient at far-infrared radiation, while drying of thicker bodies should give better results at near-infrared radiation (Nowak and Lewicki, 2004). The weight loss reduced from 87.43 to 78.26 % when the distance was increased from 5 to 10 cm (150 W and 98 min).
GA-ANN model was developed for modeling of apricot weight loss during drying with infrared dryer. In this study, ANN with 1–30 neurons was trained using GA to find the optimal network configuration. It was found that ANN with 7 neurons in one hidden layer could predict apricot weight loss with high correlation coefficient and low NMSE (0.9987, 0.0026, respectively). The prediction efficiency of the GA-ANN model for unseen data is presented in Fig. 7. The calculated correlation coefficient value for estimation of apricot weight loss show high correlation between predicted and experimental values. Table 3 illustrates the weights and bias values of optimized network, which could be applied in a computer program for estimation of apricot weight loss during drying with infrared dryer. The results showed that an acceptable agreement between the predicted and experimental data can be achieved using GA–ANN model. The results of infrared drying of apricot showed that with increasing in lamp power and decreasing in sample distance from the heat source, the drying rate was increased. The GA-ANN modeling results showed a network with 7 neurons in 1 hidden layer with using sigmoid activation function can be predicting the weight loss in apricot drying by infrared method with correlation coefficient equal to 0.9987 and mean squared error equal to 1.9215. Sensitivity analysis results by optimum ANN showed the drying time was the most sensitive factor to control the weight loss of apricot slides.

In this paper, mass transfer during the extraction of starch from potato samples in a batch extractor using distilled water was investigated. In order to have a specified geometry for modeling the problem, the samples were cut in the shape of very thin slabs. The moisture and starch concentrations were measured in different time intervals up to five hours at the solvent temperatures of 30, 45 and 55 °C, and the diffusivities of the components were estimated by fitting the solution of Fick’s second law of diffusion to the experimental data. Also 8 different experimental models were fitted to the experimental data and the kinetics of the most exact model was compared with the predictions of the analytical solution. At all stages, genetic algorithm was used for regression analysis and determining the coefficients of all equations. Comparing the values of R2, RMSE and MRE between the models, it was concluded that Page model with R2 value of 0.982, MRE value of 0.124 and RMSE value of 0.191, was the most exact model for prediction of the kinetics of mass transfer during the extraction. The estimated diffusivities were also in the range of 0.414 × 10-10 to 1.557 × 10-10.

Due to low energy efficiency and prolonged drying time of agricultural products by conventional methods, application of the new techniques such as infrared radiation, must be investigated. In this study, to drying and increased shelf life of the strawberry, infrared radiation (IR) method was used. The effect of infrared lamp power (150, 250 and 375 watts), the distance of sample from lamp (5, 7.5 and 10 cm) and time of 110 minute on drying of strawberry were examined. The results of infrared drying of strawberry showed that with increasing in lamp power and decreases in sample distance from the heat source, the drying rate was increased. With increase in infrared power from 150 to 375 watts, drying rate 69.24% increased. By increasing in sample distance from 5 to 10 cm, drying rate 23.55% decreased. Modeling of process was done with the genetic algorithm–artificial neural network (GA-ANN) method with 3 inputs (lamp power, distance and time) and 1 output (weight loss). The GA-ANN modeling results showed a network with 9 neurons in 1 hidden layer with using hyperbolic tangent activation function can be predict the weight loss in strawberry drying by infrared method (R2=0.999). Sensitivity analysis results by optimum ANN showed the infrared power was the most sensitive factor to controlling the weight loss of strawberry slides.

In this study, the thin-layer drying behavior of eggplant slices in an infrared dryer (IR) was investigated. The effect of infrared lamp power (150, 250 and 375 watt), the distance of sample from lamp (5, 10 and 15 cm), samples thickness (0.5 and 1 cm) and drying time on drying of eggplant slices were examined. The results of infrared drying of eggplant showed that with increasing in lamp power and decreases in sample distance from the heat source, the drying rate increases. With increase in infrared power from 150 to 375 watts, weight loss increased from 31.08 to 92.44%. With increase in lamp distance from 5 to 15 cm, weight loss decreased from 92.44 to 31.15%. In this study, process modeling was done with the genetic algorithm–artificial neural network (GA-ANN) method with 4 inputs (power and lamp distance, sample thickness and drying time) and 1 output for prediction of weight reduction. The GA-ANN modeling results showed a network with 14 neurons in one hidden layer with using sigmoid function can be well predict the weight loss in eggplant drying by infrared system (R=0.99). Sensitivity analysis results by optimum ANN showed the infrared power was the most sensitive factor for controlling the weight loss of samples.

Modeling the kinetics of drying of button mushroom in a hot air dryer by genetic algorithm– artificial neural network (GA-ANN) was investigated. The effects of hot air temperature at three levels 50, 60 and 70 °C, the air flow rate at three levels 1, 2 and 3 m/s on drying of button mushroom were examined. The results of hot air drying of button mushroom showed that with increasing the temperature and air velocity in hot air dryer, the drying rate increases. The increasing of the dryer temperature from 50 to 70 °C, and air velocity from 1 to 3 m/s, weight loss increased 12.2 and 12.0 %, respectively. Modeling the kinetics of thin-layer drying of button mushroom was done with the GA-ANN method with 3 inputs such as air temperature, flow rate and drying time and 1 output for predicting of weight loss. The modeling results showed a network with 16 neurons in one hidden layer with using hyperbolic tangent activation function and Levenberg–Marquardt can be well predict the weight loss in button mushroom drying by hot air (R=0.999). Sensitivity analysis results by optimum network showed that the dryer temperature was the most sensitive factor to controlling the weight loss of samples.

Banana is one of the most popular tropical fruits in all over the world with notable post-harvest losses. Due to its high moisture content preventing long preservation period. So, it needs a proper preservation method to prevent product lost especially in main produceing countries. Since banana is an un-freezable fruit, thermal processing such as drying or canning could be more appropriate for prolonging its shelf life. On the other hand, high energy consumption and being cost intensive are two most important disadvantages of thermal processing. In order to decrease the side effects of thermal process on quality parameters, pretreatment of samples could be applied to reduce time of main process.Osmotic dehydration is a non-thermal pretreatment which provides partial removal of water by immersing sample in an osmotic solution.But this process also takes a long immersion time to enough reduction of moisture. So this leads to undesirable effect on texture and colors.This study was performed to eliminate some side effects of osmotic dehydration on quality and finally introduce an optimized condition resulting best performance of process.A novel all-knowing method for optimization of process is genetic algorithm (GA) which is a search heuristic that mimics process of natural selection. It generates solutions for the optimization of problems using techniques inspired by natural evolution, such as inheritance, mutation, selection, and crossover. In this research, genetic algorithm was applied to predict optimum condition of osmotic dehydration.
Material and Osmotic dehydration was performed using aqueous solution of sucrose in concentration of 45% (w/w) for immersion time of 3 hr. The first challenge was improving mechanical properties of banana slices by adding calcium lactate to sucrose solution in concentrationsof 0, 2, 3 and 4%.For the next step in order to protect samples from enzymatic browning mixture of ascorbic acid (0.25 %) and citric acid (0, 0.5, 1, and 1.5%) were used.The pH of solution was measured for each level of adding citric acids. The efficiency of operation was estimated by computingwater loss and solid gain. Firmness of dehydrated samples wasmeasured using a texture analyzer (INSTRON, 1140, Singapore) and penetration test. Image acquisition technique was applied to measure L*, a* and b* indices.The coefficient of efficiency was defined as the ratio of water loss to solid gain and calculated to estimate performance of treatment in new condition. Finally, optimized conditionsfor maintaining the lowest solid gain and color changes, the highest water loss and firmness waterlosswere predicted by genetic algorithms method. The accuracy of model was investigated using statistical parameters such as mean absolute error (AME), normalized mean square error (NMSE),mean square error (MSE).
The results of experiments showed a significant increase of firmness by adding lactate calcium. This observation was due to complex formation between calcium and cell wall ingredients. Thesecomplexes have a decreasing effect on solid gain.Because complexes preventedmacromoleculesentering such as sucrose to the cells.On the other hand,calcium lactate and citric acid had interaction on mentioned parameters.Firmness showed less firmness when citric acid was added to the solution. Because citric acid as a chelating agents can blockdivalent cations and prevent from effective reaction with plant cells.Also citric acid can disconnect methoxyl groups from protopectinproducing softer texture.However, treated samples still showed firmer texture than control sample. It could be due to the additional effect of citric acid which makes carboxyl groups available for divalent calcium cations during conversion of protopectin to the pectin.For color parameters,only use of citric acid could not decrease the total change of color because yellow index increased due to the hydration of citric acids. But for the use of two factors, a significant decrease of total change of color was observed.For water loss, increase of solvents in each treatment led to raise of water loss due to the increase of osmotic pressure.In this circumstance determination of suitable concentration for each factorresulting best performance is complex, so it is necessary to apply a system canpredict optimized conditions. Genetic algorithms estimated optimum condition formaximum firmness and water loss, minimum solid gain and total change of color.In this condition the concentrations of lactate calcium and citric acid were %3.99 and %0.86, respectively. Also predicted values for water loss, solid gain, firmness and total change of color were earned %18.01, %5.07, 1.47 N and 11.37.MAE, NMSE and AME parameters (2.062, 0.021, and 1.099 respectively) were used for investigation of difference between estimated and experimental data which showed high efficiency of genetic algorithm for optimization of osmotic dehydration of banana.Investigating the efficiency ofcoefficient of treatments showed that application of both factors (calcium lactate and citric acid) significantly had more efficiency in comparison to the control samples regarding quality factors.

The effect of moisture content (7.1 - 28.5 %w.b.) and pea cultivars (Arman¡ Hashem and 93-93) were studied on the geometrical properties such as length¡ width¡ thickness¡ geometric and arithmetic mean diameters¡ area¡ sphericity and gravimetrical properties including 1000 kernel mass¡ volume¡ bulk and particle density and porosity. The statistic result was carrier out with completely randomized design (CRD) base on factorial arrangement with three cultivar and four moisture content (3×4). As well¡ thousand kernel mass of different pea seed cultivars were optimized basis of the function of moisture content and axial dimensions (length¡ width and thickness) by using genetic algorithm (GA) approach. The effect of cultivar on all physical properties of pea cultivars were significant¡ except thickness and porosity (P

The purpose of this study was to evaluate the effect of time, kind of extract, extract concentration and temperature on the dynamically the population of Escherichia coli (infectious agents) in the complex food system (frankfurter sausage) and Genetic Algorithm - Artificial neural network and neuro fuzzy system (CANFIS) were used for dynamic modeling of population of E.coli. At this research laboratory, Minimum Inhibitory Concentration (MIC (and Minimum Bactericidal Concentration (MBC) of aqueous and ethanolic extracts were studied using the micro broth dilution method. GA- ANN, neuro fuzzy (CANFIS) were fed with four inputs: concentration at the five level (0, 2000, 4000, 6000, 8000 ppm), kind of extract (watery, ethanol), temperature at the three level (5,15, 25 ̊С) and time (1-20). The results showed that the ANN with 1 hidden layer comprising 10 neurons, Tangent hyperbolic function, momentum training rule and percent of used data 30/30/40 for training/cross validation/testing respectively gives the best fitting with the experimental data, which made it possible to predict with high determination coefficient (R2 equal to 0.995). Also the correlation between CANFIS predictions and experimental data was very good (R2 equal to 0.96). It's worth to mention that in this research GA- ANN was better approach to simulation dynamically the population of E.coli.

Iran is the second highest apricot production in the world and studying the factors affecting the drying of this fruit and their impact is essential. Then¡ in this study the effect of osmotic solution temperature in the range 25°C to 65° C¡ at 30-120 minutes and osmotic solution concentration in the range of 30 to 60 % (w/w) on the parameters of weight reduction¡ water loss and solids gain during osmotic drying of apricots were studied. Results of osmotic drying showed that the three parameters mentioned are effective on weight reduction¡ water loss and solids gain. With increasing osmotic process time from 30 minutes to 120 minutes¡ weight reduction¡ water loss and solids gain¡ increase 21.78¡ 50.64 and 157.31 percent¡ respectively. In this study also¡ process modeling were done with the genetic algorithm–artificial neural network (GA-ANN) method with 3 inputs and 3 outputs for predictors of weight reduction¡ water loss and solids gain. The results of the modeling results using a GA-ANN showed a network with 14 neurons in the hidden layer with hyperbolic tangent activation function can be well predicted the weight reduction (R=0.98)¡ water loss (R=0.97) and solid gain (R=0.96) in osmotic drying process of apricot. Sensitivity analysis results by optimum ANN showed the osmotic solution temperature was the most sensitive factor to controlling the weight reduction¡ water loss and solids gain from apricot particle.

The goal of this study was applying the genetic algorithm–artificial neural network (GA-ANN) modeling to predict the antibacterial effect of annatto dye on Escherichia coli population in mayonnaise. Annatto has antimicrobial and antioxidant properties in foods. Annatto dye was extracted and after filtration and concentration, was dried by vacuum oven. In this study, sauce samples containing 0, 0.1, 0.2 and 0.4 percent of annatto dye were prepared and stored at 4 and 25 ˚C. Sampling and colony counting were performed during 17 days and in triplicate. In order to predict the Escherichia coli population multi-layer perceptron neural network with 3 inputs and 1 output were used. Genetic algorithm method was used to optimization number of neurons in ANN hidden layer. The results showed a network with 7 neurons in hidden layer and using a Sigmoid tangent transfer function and the Levenberg–Marquardt (LM) optimization technique and 30%-20%-50% for training/ testing/ validating process can be well predict the Escherichia coli population (r=0.999) in the presence of annatto dye. Sensitivity analysis results by optimum ANN showed the storage time as the most factor for the predicting the Escherichia coli population.

Oil uptake occurs mainly during cooling immediately after frying. Post-frying centrifuge step used to remove the attached oil to the surface of friedproduct. Response surface methodology (RSM) technique was used to develop models for the responses to the centrifugal force (100, 200 and 300 rpm) and centrifugal time (20, 50 and 80 s), frying temperature (140, 170 and 200 C) and frying time (4, 6 and 8 min), Investigated factorsoptimized with respect totheoil and moisture contents, color and textural parameters. The models showed very well fitness with experimental data. The coefficients of obtained models with RSM, was optimized by using genetical gorithms and was observedthat GA optimized models showingbetterfitnesswith theexperimentalresults than RSM models. The resultssuggest thatpost-frying centrifuge step in300 rpm for 80 sand fryingin 175-180Cfor 7-8 minleadsto French fries withlowestmoisture and oil contentand best color and texture.

Understanding the physical properties of pistachio kernel are necessary for the proper design of equipments for handling, transporting, processing, drying, sorting, grading and storage this crop. In this study, different predictive tools include artificial neural network, genetic algorithm, response surface and liner regression were used to predict mass of pistachio seed. Result demonstrated that all models have a high degree of accuracy to estimate mass of pistachio seed, so that, models of perseptron neural network, radial basis function, linear regression, linear regression coupled with genetic algorithm and response surface conjugated with neural network were able to predict pistachio seed mass with R2 value 0.9949, 0.914, 0.986, 0.995 and 0.945, respectively. Generally, comparison of the models showed that synthetic model of linear regression-genetic algorithm with R2 value equal 0.995 has the highest accuracy.