Coffee is a common drink made from roasted and ground coffee beans. The coffee plant is native to the subtropical regions of Africa and some islands in South and Southeast Asia. When the fruit of the coffee plant ripens, the coffee beans are harvested, processed and finally dried. Dried coffee beans are roasted to different degrees and depending on the desired aroma, different grades are considered for this product. Coffee is slightly acidic and can cause human irritation due to its high caffeine content. Coffee is one of the most valuable basic products, which is the second main commodity after oil. The detection of natural and unnatural impurities and additives in coffee is a constant concern, especially in relation to guaranteeing the quality of the product with the intentional or accidental addition of foreign substances that can harm the consumer, especially of an economic nature. Therefore, researchers are always trying to provide a suitable solution for detecting adulteration in coffee, which is of great importance considering the applicability of the method and obtaining the appropriate result for the tests, non-destructive and fast method. The purpose of this research was to use the olfactory machine system and artificial intelligence to detect fraud in Arabica coffee (Medium Dark). For this purpose, firstly, Arabica coffee beans from a reputable domestic company and samples of fake powders including roasted soybean powder, wheat flour Barley flour and Robusta coffee were prepared in the amount needed in the experiment. To carry out the experiments, Arabica coffee was mixed with adulterated powders with weight percentages of 10, 40, 30, 20 and 50%. For each sample of coffee and powders used for fraud, a 100% specific sample was considered. 10 grams of the mixture of each sample was added to 100 ml of boiling water and kept boiling for 2 minutes. Finally, it was kept at rest and away from heat for one minute until the particles settled and finally the supernatant was used to perform the smell test. After the step of sucking the smell of the sample by the olfactory device, the obtained data were analyzed by PCA, LDA and ANN methods. According to the obtained results, the ANN method provided a better classification than the LDA method.

   Abstract Energy, as one of the most important and necessary production factors, has a significant impact. Considering that fossil fuel resources are running out, researchers are looking to replace biodiesel fuel as a renewable biofuel with properties close to diesel. The current research seeks to model electronic nose data in predicting the oxidation stability and physical and chemical characteristics of biodiesel and to investigate the oxidation stability of biodiesel samples with the Rancimat device and to apply algorithms based on artificial and statistical intelligence. In this research, biodiesel fuels were prepared from different sources of rapeseed and sunflower oil and cooking waste with methanol and KOH (potassium hydroxide) catalyst. Each of the fuels is mixed with diesel fuel at a volume percentage of 2, 5, 10 and 20, with the help of an electronic nose system equipped with 10 sensors that can analyze the volatile components in the empty space of the sample container, physical and chemical characteristics such as density , viscosity along with their oxidation stability were analyzed by standard Rancimat method in different periods of time (each test in one month). Also, with different methods such as artificial neural network (ANN), principal component analysis (PCA), linear and quadratic discriminant analysis (LDA and QDA), support vector machine analysis (SVM) and response surface method (RSM) each month. was analyzed. The obtained results showed that for the classification and separation of pure fuels, the ANN method is able to separate pure fuels with 100% accuracy every four months. Other SVM classification methods in each four months were 94, 95, 82, 84 respectively, for QDA 100, 100, 100 and 99, in RSM these orders were 100, 93, 98 and 96 and LDA respectively in each month with accuracy 100, 98, 100 and 100% pure fuels were separated and categorized. Also, to identify and place types of pure fuels (D100, K100, WCO100, SUN100) in one group (Pure) and types of impure fuels in another group (Impure), ANN method with 100, 98.8, 100 and 100% accuracy is able to Separation of pure fuel from diesel-biodiesel combination. Keywords:

   Abstract   Dairy products, renowned for their substantial nutritional value, are pivotal in general nutrition and the food industry. However, their considerable economic significance makes them susceptible to fraudulent practices. Standard deceptive techniques include dilution with water, substitution with different types of milk, fat content alteration, and cheese infusion into milk. Incorporating whey or whey solids into dairy products is a prevalent adulteration method. The notorious 2008 incident in China, involving the illicit adulteration of milk with melamine, resulted in kidney and urinary tract complications in 294,000 children, leading to six fatalities. Deliberate melamine contamination has been observed in milk, infant formula, pet food, and other items. These incidents have spurred the development of analytical methodologies for quantifying melamine in food and animal feed. This study employs an electronic olfaction technique to discern adulterated milk powder. Experimental protocols involve preparing and combining powdered milk, whey powder, and melamine at varying adulteration levels (10%, 20%, 30%, 40%, 50%). All tests are conducted across five different whey powder and melamine concentrations, employing both dry and water-boiled testing methods with the electronic nose apparatus. Data are subjected to rigorous analysis through Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), Support Vector Machine (SVM), and Artificial Neural Network (ANN) methodologies. The PCA loading plot for dry tests reveals the substantial influence of MQ136, MQ3, and TGS2602 sensors on the principal component, whereas MQ9, MQ3, and TGS822 sensors exhibit the highest impact in detecting milk powder adulteration. LDA analysis yields an accuracy rate of 86.67% for dry tests and an impressive 95.15% for wet tests in dir="RTL" >   Keywords: Milk Powder, Adulteration, Electronic Nose, Whey Powder, Melamine, Chemometrics

   Sugar beet is one of the most suitable plants for sugar production. Sugar beet leaves produce sugar by using sunlight, and these sugars return to the root and are stored there. In general, leaves play the role of a sugar factory. The percentage of beet sugar is very important for both the farmer and the sugar factory, because the price of beet is paid based on this. The purpose of this research was to estimate the amount of sucrose in the sugar beet crop using the image processing method and to determine the best time to harvest the sugar beet crop. For this purpose, data was collected from an area of 1000 square meters of a sugar beet field, one month before the recommended harvest date and one month after that, every other day. Each time, 5 whole sugar beet plants were randomly picked and all the leaves were photographed. In order to estimate the amount of sugar beet and obtain the most suitable harvest time with the image processing method, it is necessary to have a suitable modeling between the harvest index and the harvest time of the product. The most important indicator of sugar beet harvest is sugar grade. For this purpose, using the decision tree method, we tried to select the most effective inputs from the features of the leaves and their images. Then a model was designed to find the maximum sugar level according to the harvest time using the RSM method. The best time to harvest sugar beet with the highest quality sugar was obtained on the 210th day after planting. In the tested area, sugar beet was harvested on the 215th day, which was almost consistent with the modeling calculations by the response surface. To design a decision, it is helpful to be able to inform the farmer of the approximate harvest time, the time elapsed from the day of planting to the harvest of each sample was deducted from the optimal harvest time that has the highest sugar content, and their difference is used as the output of the decision support model. The results of the decision tree showed that the average value of B corresponding to the smallest leaf area, the smallest leaf area value, and the largest leaf area value are suitable inputs for formulating a decision. For the design and modeling of decision support, three methods of response level modeling, Anfis and artificial neural networks were used. The value of R2 in the response surface method, Anfis and neural networks were 0.83, 0.832 and 0.80 respectively, and the Anfis method was selected as the best model with the highest accuracy.

   One of the most important topics in mycological sciences is the topic of identifying nonedible mushrooms and identifying them from edible mushrooms. Today, the number of people who get poisoned by consuming nonedible mushrooms is increasing. As a result, the detection and separation of edible mushrooms from nonedible ones is of great importance. Considering the research in this field and the need to perform high-cost and somewhat inaccessible tests such as GC, methods should be sought. It was an alternative to these tests. The method of artificial intelligence and electronic nose is a non-destructive and accessible method with a much lower cost compared to GC. In fact, the purpose of this experiment is to study an olfactory system to distinguish edible mushrooms from poisonous mushrooms based on electronic nose technology and to help maintain health and treat diseases and reduce poisoning caused by poisonous mushrooms. Therefore, in this research, an olfactory system was used to detect and identify edible and nonedible mushrooms. ANN methods, PCA principal component analysis, LDA linear discriminant analysis, QDA 2nd degree linear discriminant analysis and SVM support vector machine were used to analyze the data obtained from the olfactory system. The classification of the data obtained from the signals obtained from the sensor array showed that the LDA, QDA and ANN methods have a very good performance in separating mushrooms based on their edible and nonedible nature and high accuracy in classification. Obtained. The use of QDA method to separate and classify different types of edible and nonedible mushrooms was more effective and accurate than LDA method.

  Tomato paste is the main tomato product that is prepared commercially or traditionally at home. According to the definition of the World Health Organization, adding any unauthorized and harmful substances to the food basket of humans and animals is called fraud. Sometimes the profit seeker puts in food to reduce the production costs and get more profit and endangers human health. Therefore, it is necessary to use new methods with a fast and high response level to identify additives in the paste. It is completely reduced or destroyed, in this case such a food item is corrupted and it is called rotten. In this research, it is tried to be able to provide the nutritional health of the communities with the help of smelling machine, in addition to detecting corruption in tomato paste, by identifying common frauds, including the identification of unauthorized preservative additives in tomato paste. In this research, the work was carried out in two sections, examining the experiments of the fraud section and the experiments of the corruption section. In order to evaluate these changes, the physicochemical tests of the paste, which included the measurement of pH, acidity, Brix, sediment weight percentage and smell test, were performed. According to the results of the confusion matrix of LDA and C-SVM methods, both methods performed well in detecting different percentages of potassium sorbate and sodium benzoate. LDA also performed better than C-SVM with 100% recognition accuracy. C-SVM and PCA could distinguish with high accuracy the samples containing sodium benzoate and potassium sorbate in 0.1 and 0.05 percent. The PLS model was the best model for predicting acidity and the MLR model was the best model for predicting sediment weight percentage, Brix and pH. TGS2620 sensors,. MQ135, TGS2602 had the highest sensitivity in identifying adulterated tomato paste samples.

Due to important medicinal properties, Garlic (Allium Sativum) known as a strong medicine, and cultivated in many regions of the world. The high sensitivity of this product after peeling, such as changes in color, shape, texture, and appearance, as well as losses of moisture, is problematic. The shelf life of garlic at ambient, refrigerator temperature (4 °C) and freezer (-18 °C) is 21 and 35 days, respectively. There are various methods to prolong the shelf life of food and agricultural products during the storage time, Which the use of Nano packaging films, modified atmosphere packaging and the use of edible coatings can be mentioned. In this research 2 types of packaging films (common and Nano film), 2 types of atmosphere (ambient and MAP) were used. Modified atmosphere packaging (MAP) was included 5 % CO2, 1 % O2 and 94 % N2. The harvested Garlic was peeled manually with high precision. Then stored at 3 temperatures (25, 4 and -18 °C) for 21 days and 2 temperatures (4 and -18 °C) for 35 days.   hysiological properties were evaluated by Spectroscopy method. The POD, SOD, Catalase and protein content were measured weekly. The statistical analysis was investigated in the form of a completely randomized design with a factorial test. The results showed the general increasing in POD and Protein content at 4 C and decreasing trend at 25 C. increasing trend of SOD at 4 C and decreasing at 25 C, increasing trend of catalase at -18 C and decreasing at 4 C were observed. The effect of storage time (during 21 and 35 days) was significant (at 1 %) on all parameters except POD. The effect of temperature (during 21-day period) was significant on Catalase (at 1% level) and SOD (at 5 % level), and was insignificant on POD. The results showed that temperature was insignificant on all parameters during 35-day period. The effect of packaging film was significant on protein content (at 1 % level) and POD (at 5% level) during 21-day period and was insignificant on SOD and Catalase (during 21-day period), Also was insignificant on all parameter except SOD during 35-day period. Finally, the result showed packaging atmosphere had not significant effect on properties.   

In order to maintain the quality of fruits and increase their shelf life, extensive research has been done on packaging methods, especially the use of nanomaterials in packaging. Due to the high medicinal and nutritional properties of garlic and also the sensitivity of its storage period after peeling, the packaging of this product is of great importance. In this study, garlic was packaged in two normal and nano films at temperatures of 25, 4   and -18   ° C and three modes of normal atmosphere, vacuum and modified atmosphere. Measured properties include mechanical properties (Fmax, Emod and deformation percentage), chemical properties (PH and TSS), colorimetric properties (L *, color change and browning index) as well as the amount of gas (O2 and CO2) inside It was packages. Data analysis was performed in two sections of 14 days including all three temperatures and 35 days including refrigerator temperature and freezer temperature. Data were analyzed by statistical methods and artificial neural network (ANN). The trend of changes during the storage period in mechanical properties (except deformation), the amount of CO2, TSS and L * decreased and in pH, the percentage of deformation, color change and browning index were reported as increasing. The results of statistical analysis showed that in the 14-day period, temperature changes had a significant effect on the measured parameters and in the 35-day period, temperature and the interaction of temperature and atmosphere had a significant effect on all parameters. The least changes in the measured properties occurred mainly in the nano film. In neural network (ANN) analysis, the output of the best model for the effect of treatments on properties, validation performance diagram, data regression coefficient (experimental, training and general) as well as data regression line fitting was measured.   

مطالعه­ي مورد نظر جهت بررسي اثر روش تجويز مكمل‌هاي سلنيوم و ويتامين اي در اواخر آبستني بر غلظت سلنيوم، مس، روي و آهن خون و شير ميش‌ها و خون بره­هاي حاصل ازآن­ها انجام شد. در اين مطالعه 18 راس ميش آبستن از گله 50 راسي آبستن به صورت تصادفي انتخاب و در 3   گروه 6 راسي به صورت انفزادي نگهداري شدند. تيمارها شامل:تيماراول ،شاهد (دريافت مقدار 10 ميلي‌ليتر مكمل سلنيوم و ويتامين اي به صورت تزريقي دو هفته قبل از زايش؛ هر ميلي ليتر حاوي 5/0 ميلي‌گرم سلنيت سديم و 50 ميلي گرم ويتامين اي). تيمار دوم، دريافت مكمل سلنيوم و ويتامين اي بصورت خوراكي   3/0 ميلي‌گرم سلنيوم و 50 ميلي‌گرم ويتامين اي در كيلوگرم ماده خشك مصرفي ) مخلوط با جيره و به‌صورت روزانه از چهار هفته قبل از زايش و تيمار سوم تجويز 10 ميلي‌ليتر مكمل سلنيوم و ويتامين اي به‌صورت تزريقي چهار هفته قبل از زايش (5 ميلي‌ليتر) و دو هفته قبل از زايش (5 ميلي‌ليتر)بودند. ميش‌ها در روزهاي اول، وسط دوره   و قبل از زايش توزين   و امتياز بدني آنها نيز تعيين شد.از رگ گردني ميش­ها   قبل از تجويز مكمل و زمان زايمان خون‌گيري شد. از بره­هاي تازه متولد‌شده نيز قبل از مصرف كلستروم   (آغوز) و روز 14 بعد از تولد از طريق رگ گردني خونگيري شد. نمونه هايي ازآغوز و شير جهت اندازه­گيري غلظت سلنيوم، مس، روي و آهن برداشت شد. نتايج آزمايش نشان داد كه ميانگين وزن ميش­ها در   4 هفته قبل از زايش تا زمان زايش، تفاوت معني‌داري نداشتند ولي در دو هفته بعد از زايش وزن ميش­ها در گروه تيمار خوراكي نسبت به تيمار شاهد بيشتر بود(05/0>P).اختلاف معني‌داري در امتياز وضعيت بدني بين تيمارها مشاهده نشد. بعد زايش غلظت سلنيوم پلاسماي ميش‌ها در تيمار خوراكي بالاتر از گروه‌هاي ديگر بود (05/0>P). در روز تولد ميانگين غلظت سلنيوم پلاسماي خون بره‌ها در تيمار خوراكي   بالاتر از دو گروه ديگر بود (05/0>P) اما در دو هفتگي بين بره‌هاي تيمارهاي مختلف تفاوتي از لحاظ اين فراسنجه مشاهده نشد (05/0>P). غلظت سلنيوم در آغوز ميش­هاي گروه دريافت‌ كننده مكمل خوراكي نسبت به تيمار شاهد افزايش يافت(05/0>P)اما تفاوت معني­داري بين تيمارها از لحاظ غلظت سلنيوم در نمونه هاي شير   مشاهده نشد (05/0P>). تفاوت معني­داري بين تيمار‌ها از لحاظ غلظت‌هاي آهن، مس و روي پلاسماي خون، آغوز و شير ميش‌ها وجود نداشت (05/0P>) . تفاوت معني‌داري در غلظت‌هاي آهن، مس و روي پلاسماي خون بره­ها در زمان تولد بين تيمارها وجود نداشت (05/0P>) اما در دو هفتگي بره­ها، غلظت مس پلاسماي خون در تيمار خوراكي نسبت به تيمار شاهد بالاتر بود (05/0>P). در نتيجه گيري كلي، غلظت سلنيوم خون ميش‌ها و بره‌هاي حاصل از آنها و غلظت سلنيوم آغوز در تيمار خوراكي بيشتر بود(05/0>P)اما به غير از مس در بره ها، اختلاف معني‌داري در غلظت عناصر آهن، روي و مس مشاهده نشد. با توجه به نتايج اين مطالعه، بهتر است از مكمل سلنيوم و ويتامين اي به صورت خوراكي استفاده شود

   Vegetables and summer vegetables play an important role in human health due to their high fiber and antioxidant properties. Green cucumber is one of the oldest cultivated vegetables and has a known history of more than five thousand years. Green cucumber with the scientific name of Cucumis Sativus belongs to the squash family Cucurbiteacae, which is one of the most important plant families and includes 90 genera and 750 species. Researchers have concluded that the use of urea fertilizer will increase the yield of vegetables, which has led to the indiscriminate use of urea fertilizer by farmers. The use of urea fertilizer on farms should be controlled because excessive use will not only increase yield but also cause nitrate accumulation. Due to the fact that vegetables and summer vegetables have the ability to absorb and retain large amounts of nitrite and nitrate, so the consumption of such products by humans can endanger health. For this purpose, criteria have been considered that label products with authorized consumption of pesticides and chemical fertilizers as a healthy product. Detection of urea fertilizer overdose in farms is done using existing technologies such as chromatography (GC) or spectrometer gas chromatography (GC / MS) which is very costly and time consuming and requires specialized users. has it. Therefore, it is necessary to look for an easy and low-cost solution that can perform the test in the shortest time. In this study, five levels of urea fertilizer in cucumber were classified using electronic nasal method and chemical analysis and by chemometric methods. Urea fertilizer levels were zero, 100, 200, 300 and 400 kg / ha. In each urea fertilizer level, two harvests were performed at intervals of four and five months after sowing. Electronic nose technology is a modern and advanced technology that has many applications in the agricultural industry. In this study, an olfactory machine with eight metal oxide semiconductor sensors was used to detect the amount of urea fertilizer used in green cucumber cultivation, and Kojeldal was used to measure phosphorus by spectroscopy, flame diffusion potassium and nitrogen. Odor machine data were analyzed and classified by ANN, SVM, LDA and QDA methods and chemical analysis by PCR, PLS and MLR methods.

In recent years, the growing population and increasing demand has led to an increase in food adulteration by profiteers. Adulteration in food products, in addition to reducing product quality and financial losses, is also detrimental to the health of consumers, so it has caused concern and weakened consumer confidence. Vinegar is a solution of acetic acid and other chemicals, such as flavorings, produced from the fermentation of various fruits. Vinegar has an important role in human life and health due to its medicinal properties, cleansing, preparation of pickles and use in various foods as a condiment. Mixing natural vinegar with relatively cheaper natural vinegar, white market vinegar, water and using industrial acetic acid in the production of artificial vinegar are the most common methods of adulteration in the market. It is difficult to find a technique that can easily and reliably evaluate the quality parameters of vinegar. Therefore, in recent decades, researchers have turned to the use of visual, olfactory, taste and computer methods in the food industry. Electronic nose is one of the new methods that has recently been highly regarded by researchers in agriculture and food industry, especially in the field of food quality assessment. In this study, a portable system was developed and implemented to evaluate the detection of adulteration in two types of natural vinegar (grape and apple). The electronic nose system consisted of eight metal oxide semiconductor sensors. Counterfeit vinegar was prepared by mixing grape and apple vinegar and both with white market vinegar, acetic acid and water in different proportions. The features extracted from the signals obtained from the system were processed by principal component analysis (PCA), artificial neural network (ANN), linear resolution analysis (LDA) and quadratic linear resolution analysis (QDA). In the main component analysis, simultaneous comparison (17 groups) of grape vinegar with a total variance of 88% separation was performed and based on the loading diagram, TGS2620 and MQ136 sensors were introduced as the best sensors in the 0.999, MSE = 0.000136. Apple cider vinegar was 0.998, MSE = 0.000513. According to the titration results, it was proved that in PCA analysis, the acetic acid level did not affect the variance of the samples chewed with grape vinegar and the lower the level of acetic acid in the chewed samples with apple cider vinegar, the higher the percentage of variance between samples. According to LDA and QDA analyzes, the smaller the difference between the acetic acid level of vinegar and the substance with which it is mixed, the weaker its detection and >  

   Abstract Vegetable oils, whose fatty acid hydrocarbon chain structure is very similar to the hydrocarbon structure of petroleum products, are an ideal alternative to petroleum lubricants due to their properties such as renewability and biodegradability. Biolubricants are obtained by modifying the chemical structure of vegetable oils. In this research, a two-stage transesterification-polyol ester method was used to produce a biolubricant based on Camellina oil. The required heat energy for the reaction was supplied by a microwave with a mechanical stirrer. A vacuum pump connected to a condenser was used to complete the transesterification reaction. Methanol condensing and returning it to the reaction vessel. In the initial stage, Camellia oil reacts with methanol in the presence of KOH catalyst and the microwave device with a constant power of 800 watts was set to Medium mode (12 seconds on and 16 seconds off) and stirring is performed continuously. Methyl ester, trimethylpropane (TMP) alcohol and potassium carbonate catalyst (K2CO3) were used in the biolubricant process. The microwave was set to Low (5 seconds on and 30 seconds off) and stirring is done continuously. Nuclear-hydrogen magnetic resonance spectroscopy (H-NMR) was used to analyze the efficiency of the transesterification reaction. In the biodiesel production stage, time factors, catalyst weight percentage and molar ratio of methanol to oil, and in the biolubricant production stage, factors such as time, catalyst weight percentage and biodiesel molar ratio to trimethylol propane (TMP) were investigated using RSM method and Box Behnken experimental design. . For biodiesel reaction, the regression model was quadratic and had an coefficient of determination (R2), standard deviation (Std. Dev.) And coefficient of variation (C.V.) of 97.80%, 3.02, and 4.03%, respectively. The optimization results showed that the highest biodiesel yield of 95.31%   was achived at the reaction time of 5.85 min, catalyst concentration of 1.26%, alcohol to Camellina oil molar ratio of 6.91, and desirability of 0.95. For biolubricant reaction, the regression model was quadratic and had determination coefficient of 97.97%, standard deviation of 0.91 and coefficient of variation of 1%. The optimization results showed that the highest biolubricant efficiency was obtained in 67.8 min, catalyst concentration of 1.4 wt% and molar ratio of 3.5. Under such conditions, the reaction efficiency was 94.3% with a desirability of 0.975. Some properties of produced biodiesel and biolubricant were evaluated and compared with EN14214, ASTM D6751, ISO VG10 and ISO VG22 standards.

  تجزيه گرمايي زيست توده در محيطبدون اكسيژن يا با اكسيژن اندك را گرماكافت مي‌نامند كه محصول اين فرآيند دي اكسيدكربن، گازهاي سوختي، بخار قيري و جزء جامدي به نام بيوچار است. فرآيند گرماكافتراهي براي تبديل زيست توده به مواد با ارزشتر نظير بيوچار است. بيوچار ماده ايجامد و داراي محتواي كربن بالاست كه رايج ترين مورد استفاده آن در كشاورزي بهعنوان اصلاح كننده خاك است. محققان در سال هاي گذشته تاثير استفاده از بيوچار برخصوصيات فيزيكي وشيميايي خاك را مورد مطالعه قرار داده اند و مشخص شده است كهافزودن بيوچار به خاك كيفيت خاك را بهبود مي‌بخشد. خصوصيات فيزيكي و شيميايي بيوچارتحت تأثير عوامل مختلفي از جمله نوع مواد اوليه، شرايط واحد گرماكافت، سرعت گرمادهي،مدت زمان گرماكافت وللفعوامل متعدد ديگري قرار مي‌گيرد . دامنه گسترده فرآيندگرماكافت منجر به توليد بيوچارهايي   كه ازنظر خواص شيميايي و فيزيكي مختلفي نظير تركيب عنصري و خاكستر، وزن مخصوص، تخلخل،توزيع اندازه منافذ، سطح ويژه، pH، جذب و دفع آب و يون ها و بسياري خواص ديگر متفاوت هستند مي‌شود. هدفاز اين مطالعه، بررسي اثر تغيير دبي هوا و دماي محفظه در گرماكافت اكسايشي بسترثابت بر روي عملكرد بيوچار، محتواي خاكستر، وزن مخصوص و pH بود. بدين منظور يك دستگاه توليد بيوچار اكسايشي بسترثابت با قابليت تغيير در دماي محفظه و دبي هواي خروجي طراحي و ساخته شد.آزمايش‌هادر چهار دبي هواي 20، 25، 30 و 35 ليتر در دقيقه ونيز چهار دماي 350، 400، 450 و500 درجه سانتي‌گراد براي كاه و كلش گندم انجام شد. نتايج نشان داد كه افزايش دبيهواي   خروجي   از محفظه و افزايش دماي محفظه، سبب افزايش ميزانخاكستر وpH شد. درحالي كه تغيير اين پارامترها سبب كاهش وزن مخصوص ظاهري و عملكرد   بيوچارتوليدي شدند.     

   Abstract Pomegranate, known as Punica granatum L., belongs to the Punicaseae family. Iran holds the 60,000 hectares of land under cultivation and production of 800,000 tonnes, Iran is the world's first pomegranate producer. Pomegranate seeds can be made from water, grenadine, potion, syrup, jam, jelly and so on. Healthy and desirable food quality is currently playing an important role in the food industry.   Adulteration of fraud in the food industry has always challenged the scientific community. Hence, attention has been method on the use, smell, and taste and computer technology in the food industry over the last few decades. Electronic Nose is a new method that has recently been considered by researchers adulterated in pure grenadine. he response characteristics of the sensors to the volatile compounds of the samples were extracted and used as inputs to the pattern recognition model. 30 grams were tested for each sample. According to the results obtained for the mixture of pure grenadine with grape syrup, Palm sap 92 and 94% of variance by PCA method, in order to >      Key Words: Adulterant, Electronic Nose, Foodstuffs, grenadine, sensors.

   Food fraud has increased dramatically in recent years. In addition to affecting product quality and causing financial losses, food fraud also has adverse effects on consumer health. And it has raised a lot of concerns about food consumption. One of the most widely used foods in the food industry are edible oils. Which is used in cooking. Solid vegetable oils, or oils such as palm oil, are high in saturated fatty acids, which can cause high blood fats, high cholesterol, and eventually clogged arteries. Cardiovascular از are the most important causes of death. The oils used in the food basket of people should have at least 2% of saturated and trans fatty acids. One of the reasons for usingPalm oil is cheaper than other sesame, corn, soybean and sunflower oils, and due to higher prices of other oils than palm oil, the share of imports of this oil compared to other oils has increased in the last two years. In cooking, a variety of oils are used, each taken from a separate source From grains such as corn to fruits such as olives, nuts such as walnuts, almonds and hazelnuts, and seeds such as canola, safflower and sunflower. It is difficult to find a way to easily determine the quality of edible oils. Therefore, in recent years, olfactory, visual, taste and computer methods have been used more in the food industry. Electric nose is a new method that is considered in agriculture and food industry today In this study, a system has been used to detect and estimate the amount of palm in edible oils. The optoelectronic system consisted of ten metal oxide semiconductor sensors. During the test, the voltage response of the sensors at a certain schedule to inject the head space into the sensor housing and then clean the housing in the next step was done automatically and collected by the data system. Types of edible oils were prepared by mixing with palm oil in different proportions. The properties extracted from the signals obtained from the electrical nose were processed by principal component analysis and artificial neural network methods. Based on the results, the olfactory machine is able to detect palm oil   

   مقدمه: انرژي به‌عنوان يكي از مهم­ترين و ضروري ترين عوامل توليد ، داراي اثرات قابل‌توجهي است. با توجه به اين‌كه منابع   سوخت هاي فسيلي رو به اتمام است، پژوهشگران به دنبال جايگزين كردن سوخت بيوديزل كه يك سوخت زيستي قابل‌تجديد و داراي خواصي نزديك به گازوئيل است، مي­باشند. در آزمايش هاي مربوط به توليد بيوديزل و در كنترل كيفيت اين محصول از دستگاه GC استفاده مي­گردد و درصد بيوديزل توليدي را اندازه­گيري مي نمايند. استفاده از دستگاه GC ، به دليل هزينه بسيار بالاي آن در همه جا امكان پذير نيست لذا در اين تحقيق تلاش مي­شود درصد بيوديزل توليدي را با استفاده از بيني الكتريكي و تكنيك­هاي هوش مصنوعي تخمين زد. روغن انتخابي در اين تحقيق روغن پسماند به همراه روغن گياه رزماري مي باشد. اهداف: 1. توليد بيوديزل از روغن پسماند خوراكي و روغن گياه رزماري بوسيله ي واكنش ترانس استريفيكاسيون    2. تخمين درصد بيوديزل توليد شده با استفاده از بيني الكتريكي و تكنيك هاي هوش مصنوعي روش تحقيق: در ابتدا از روغن پسماند خوراكي و روغن گياه رزماري با استفاده از واكنش ترانس استريفيكاسيون و متانول ، بيوديزل توليد مي­شود. روغن پسماند خوراكي از آشپزخانه ها جمع آوري شده و پس از فيلتر شدن ، ناخالصي­هاي آن گرفته مي­شود و روغن گياه رزماري به آن افزوده مي شود . براي توليد بيوديزل پارامترهايي ازجمله مدت زمان همزني، مقدار متانول مصرفي ، سرعت همزني ، مدت زمان سيركوله و مقدار روغن رزماري مصرف شده پس از توليد بيوديزل، متغير و پارامترهايي از جمله   نوع كاتاليزور و نوع الكل ثابت در نظر گرفته   مي­شوند. الكل مورد استفاده در آزمايش حاضر متانول و كاتاليزگر KOH مي­باشد.    جدول 1 ماتريس آزمايش‌هاي توليد بيوديزل سطح 3 سطح2 سطح1 عامل 8:1 6:1 4:1 مقدار متانول مصرفي (مولي) 20 10 2 مدت زمان همزني 1200 900 600 سرعت همزني(دور بر دقيقه ) 9 6 3 مدت زمان سير كوله   (دقيقه ) 5/2 25/1 5/0 مقدار روغن رزماري مصرف شده (گرم)    براي پرهيز از اشتباه و به حداقل رساندن خطا در اندازه‌گيري‌ها مبناي محاسبات جرم در نظر گرفته مي­شود و با استفاده از ترازوي ديجيتال با دقت 01/0 گرم، مواد وزن خواهند شد. براي توليد بيوديزل مقدار 500 گرم از روغن را به ورودي   رآكتور مي­ريزيم و پس از تهيه­ي محلول پتاسيم متواكسيد با توجه به نسبت مولي الكل به روغن   ، محلول به روغن درون رآكتور اضافه مي­شود و آزمايش با شرايط نشان داده شده در جدول 1 توسط رآكتور انجام خواهند شد . پس از اتمام مراحل توليد بيوديزل ، روغن رزماري با نسبت­هاي مختلف به بيوديزل اضافه و از هر كدام از آزمايش ها نمونه گرفته خواهد شد .

,urea fertilizer,electronic nose,artificial neural net work ,basil,summer

   The present study seeks to identify and classify the biodiesel from various oils and alcohols using the olfactory machine technique and employing artificial intelligence and statistical algorithms. In this research, biodiesel fuels were prepared from different sources such as rapeseed oil-methanol (MK), corn oil-methanol (MZ), rapeseed oil-ethanol (EK), corn oil-ethanol (EZ) and combined fuel (EK & MZ). Each of these fuels were mixed in volume percentage of 2, 5, 10, 20 and 80 with diesel fuel. The data were collected with help of an electronic nose system equiped 8 metal oxide semiconductor sensors. The normalized data were analyzed by various methods such as artificial neural network (ANN), principal component analysis (PCA), linear and quadratic discriminant  analysis (LDA and QDA), support-vector machine (SVM) and response surface methodology (RSM). The results showed that ANN was able to classify pure fuels with a precision of 100%. Other classifier   methods, QDA, SVM, RSM and LDA, were categorized pure fuels with accuracy of 94.4, 93.3, 92.2 and 86.7 percent, respectively. Also, ANN method was able to identify and classify any pure fuels (MK100, MZ100, EK100, EZ100, EK & MZ100, G) in one group (Pure) and various disel-bidisel blends (B2, B5, B10, B20, B80) in the other group (Impure).   ANN and LDA were more powerful methods than other for idenfying   the fuels of B2, B5 and B20. The classification accuracy of both methods for B20 was 100%. For discriminant   of B5 and B2, the ANN method had an accuracy of 98.7%, while the LDA method had a precision of 100% and 97.3% respectively. By averaging the performance parameters of various models for the categories used in this study can be said that the ANN model had better performance with an average precision, sensitivity and specificity of 98.5, 98.8 and 99.5 percent than other models.

   Abstract   Iran is one of the most susceptible regions of citrus cultivation, which had many advances in recent decades in the order cultivation and production of citrus, so that cultivated area and its amount annual production, of our country determined agreement among the top 10 countries in the world. in this research, after photography from Tangerines (ripe, half-ripe and unripe) using camera canon Pc1339-12.1 mega pixels and preprocessing operations, segmentation and >   Keywords: Citrus, >

The energy issue is one of the most importantprobmlems considered globally. Fossil fuels are the largest energy sourcecurrently. Increasing greenhouse gas emissions has led to climate change invarious climates. Increasing the environmental pollution caused by the use ofthese energies has led various communities and corporations to move towards renewableenergy sources. One of the alternative solutions for fossil fuels is using ofbiodiesel fuel. The major problem with the use of biodiesel is the low powerand torque generated compared to pure diesel. Therefore, the use of additivesto biodiesel is proposed to overcome this problem. In the present study, thediesel-biodiesel blens, nano-aditives in the presence of magnetic field wereused for improving the engine performance and reducing the emissions. The wastekitchen oil was used as source of biodiesel. The biodiesel ratio in fuelmixture is 0, 5 and 10 volume percentage of diesel. Also, used Nanomaterialswere cobalt and cerium nano-oxide. Two neodymium magnets - 42 grade were usedto affect the fuel. The magnetic field was placed on the fuel line and the fuelwas exposed to the magnetic field before entering the engine. In this study,the effect of six parameters as magnetic fields (0, 225 and 4500 Gauss),biodiesel (0, 5 and 10% volume of diesel), nano aditives of Cerium oxide (CeO2)and or Cobalt oxide (Co3O4) with concentration of 0, 20and 40 ppm, ratio of nanomaterials (Cerium oxide to Cobalt oxide (0, 50 and 100%), engine speed of 1200,1800 and 2400 rpm, and engine loads of 25, 50 and 75%using Box-Behnken experiment design and RSM method. Optimization of theparameters of the engine performance and emissions index indicate that the bestparameters were magnetic field of 1561.66 Gauss, concentration of nanomaterialsof 12.25 ppm, nanomaterials ratio of 56.37%, biodiesel ratio of 4.97, enginespeed of 1962 rpm and engine load of 16.14%. In these conditions, engine torqueof 67/14 Nm, engine power of 36.3 kW, brake special fuel consumption (BSFC) of2772.5 gr / kW.h, carbon monoxide (CO) of 135.0% vol., Carbon dioxide(CO2)of0.27%, unburned hydrocarbons (UHC) of 0.032 ppm and nitrogen oxide (NOx) of 4ppm with a desirability of 0.89 were obtained.  

  Today, due to the increasing population of the earth, the need for food supplies has increased. On the other hand, due to the reduction of water resources, water management needs, especially in the agricultural sector, are felt. For this purpose, several methods have been used to reduce water consumption, one of the most modern ones being precise irrigation, namely, the field moisture map. In this research, we attempted to extract from a sorghum field to a half-hectare wetland map by multi-capture and image processing. For this purpose, initial experiments were carried out to obtain the best height and best day of the day to be photographed. The farm was then divided into small pieces. Captured from each piece by a multi-capture. From different points of the farm, the samples were taken by recording the coordinates of the spots to obtain the percentage of moisture content of the plant in those points. Finally, the images were aggregated by the Arc GIS software Then the MATLAB software performed various image processing operations on the image. These operations include: plotting, separating the field, finding the neighborhood of points, extracting different data from neighboring points, modeling effective input parameters to estimate the plant's moisture content, applying model output to segmented images, aggregation, and so on. The best model obtained by neural networks consists of four roughness inputs, the mean values of the CM channel channel monochrome channel M, the mean values of the Y single channel channel of the CMY channel, the mean values of the channel B of the RGB channel, and the output of that percentage The weight of the plant was R2 and MSE of the model was 885/0 and 0/004, respectively. In the case of classification of the plant's moisture content into five classes of very watery, loose, moderate, low water and very low water, the perturbation matrix of the model was 90.5%.

  (According to the Environmental Bureau, the rapid growth of the city and industry and the growing population growth has causedThat the amount of atmospheric pollutants has increased more and the suspended particles become the most important air pollution in metropolitan cities.Air pollution causes heart and lung disease.UHC and CO emissions are caused by incomplete combustion of fuel and air mixtures,While NOx and CO2 emissions are generated by the engine in the engine,Also, fine and tiny particles suspended from coarse particles are more dangerous.The purpose of this study was to use various ratios of bioethanol fuel with gasoline in different engine periods in the state of the EGR and to use different ratios of bioethanol fuel with gasoline and different temperatures of the intake air to the engine in different engine periods in the EGR mode with the aim Reducing the amount of pollutants and particulates in both the EGR and the EGR mode in the petrol engine pride.In this research, the Pride engine was used with the Siemens fuel system.For testing, fuel mixtures with bioethanol and conventional gasoline were used, fuel with ratios of 0, 5 and 10% bio-ethanol was used.The experiments were carried out at three levels of inlet air temperature of 12, 15 and 18 degrees Celsius and 1000, 2000 and 3000 rpm.Each experiment was repeated three times.The Airrex Five Gas Testing Machine was used to measure the contamination.Particles suspended by the IAQ 3016 particle counterpart model manufactured in the United States were measured.The effect of fuel type treatments and engine speed on exhaust and particulate exhaust gases in the EGR state and the effect of fuel type treatments, inlet air temperature and engine speed on exhaust and particulate emissions in EGR mode were investigated.Data were analyzed by    software and factorial experiment in a completely randomized design with comparisons of the meanings using Duncans multiple range test.The effect of engine speed and fuel type and the interaction between engine speed and fuel type in the EGR state on all traits were significant at 1% level.And the only interaction between the engine speed and fuel type on carbon dioxide (CO 2) has had a significant effect on the 5% level.Also, the effect of engine speed on particles of 5 microns and the effect of fuel type on particles of 3 microns is also significant at 5% level.In EGR mode, the effect of engine speed on the temperature of the exhaust air and contaminants of carbon monoxide (CO), carbon dioxide (CO2), uncured hydrocarbons (uhc), nitrogen oxides (NOx) and suspended particles at 1% have been.The effect of fuel type on air temperature was not significantBut on other traits it is significant at 1% level except for 0.3 micron particles which is significant at 5% level.Also, the effect of air temperature on all studied traits is significant at 1% level and 0.3 ?m for suspended particles at 5% level.According to the results of Duncans test, with increasing engine speed, the amount of carbon monoxide (CO) and unburned hydrocarbons (uhc) contaminated, but the amount of carbon dioxide (CO 2), nitrogen oxides (NOx) and suspended particles increased .By increasing the percentage of bioethanol, the amount of carbon monoxide (CO), carbon dioxide (CO 2), unreacted hydrocarbons (uhc) and suspended particles decreased, but the amount of nitrogen oxides (NOx) increased.For the effect of engine speed and fuel type in EGR, the amount of pollutant emissions was lower than that of the EGR, but the particle size of 0.3 micron in the EGR state was less than that of the EGR.

Environmental contamination due to the combustion of fossil fuels and the ending of these resources, as well as the increase of pathogens, has attracted the attention of many governments and researchers to wards replacing and using more efficiency clean fuels to reduce pollution. Biodiesel, has long been considered as one of the major fuels. This kind of fuel, which is a fatty starch, is obtained from oil sources such as vegetable and animal oils during transesterification catalytic process on these oils and fats. Oilseeds plants have high water requirements, but Camelina is also cultivated in drought and rainfed conditions. In this study, extracted Camelina oil via cold- pressing was used for the production of biodiesel. MgO / Fe3O4 @ SiO2 Nano catalyst was synthesized by co-sediment method. Then, this Nano-catalyst was used in the transesterification reaction to production of biodiesel. Different variables were effective on both the transesterification reaction and the performance of MgO / Fe3O4 @ SiO2 nano-catalyst. The main purpose of this research was to identify and optimize these variables in order to achieve the maximum amount of the biodiesel production. These variables were including of calcinations temperature, calcinations time, and the weight percent of the active phase to the base. The reaction temperature of 70?C, reaction time of 5hours, molar ratio of alcohol to oil of 18:1, and the weight percent of catalyst to oil of 3%(w/w) were considered as the operating conditions for transesterification reaction. Finally, the calcinations temperature of 650?C, calcinations time of 3 hours and the weight percent of the active phase to the base of 55%(w/w) were reported as optimization conditions. The efficiency of the biodiesel production of 99% was obtained at optimization conditions using the transesterification reaction and in the presence of the MgO / Fe3O4 @ SiO2 nano-catalyst. The characterization of the MgO / Fe3O4 @ SiO2 nano-catalyst using different techniques such as XRD, SEM and FT-IR was determined.

Using efficient policies in the operation of reservoirs has become very important due to the occurrence of periodic droughts and also limitation of water resources in Iran. The management of reservoirs is one of the most effective non-structural ways to overcome of these limitations such as water resources scarcity, water demands increasing and finally occurrence of drought. Using optimization techniques in optimal operation of water resource systems are one of the solutions that can reduce the effects of water shortages. In this study, optimal operation of Jamishan reservoir with the aim of optimal water allocation from reservoir is considered to supply the agriculture water demands of Dinavar and Chamchmal plains using stochastic and deterministic dynamic programing approaches. Applying of the water allocation priorities between release and storage targets to supply water demands in drought conditions is considered. In this study a 41-year hydrological period (from 1971 - 1972 to 2011-2012 years) has been used. Different interval of reservoir storage (3, 5, 7 and 10) and reservoir inflow (3, 5, and 7) were used for dynamic discretization of stochastic dynamic programming (SDP) model. The best reservoir and inflow and I=3 respectively. Stochastic dynamic programming was applied to Jamishan reservoir with these classes in seasonal and monthly period by several objective functions. Water allocation results and reservoir rule curve have been presented for each period. In case, 7 interval classes for reservoir storage and average monthly and seasonal reservoir inflow in each period were applied in dynamic programming (DP) model with several objective functions. The water allocation results were compared with SDP model. The results confirm that the SDP model had the better performance rather than the DP model in water allocation and reservoir rule curve with the least objective function.

In recent years, the rate of food fraud has increased significantly. Fraud in food products, in addition to affecting product quality and financial losses, has a negative impact on consumer health, thus much concern about the consumption of food products is needed. Cow ghee is one of the main souvenirs of Kermanshah city and is also used in cooking. Mixing cows ghee with vegetable oils, animal fat and margarine are common ways of cheating of this product. It is hard to find a technique that can easily and reliably measure the quality of the cow ghee. Now days, methods which use smell and taste technics are developed. Electrical nose is a new method that is used in agriculture and food industries, especially in the field of quality food research. In this research, a portable system was developed and implemented to evaluate the detection of adulteration in pure cow ghee. The electronic nose system was constructed based on eight metal oxide semiconductor sensors. During the tests, the voltage response of the sensors was collected by the data acquisition system. Different types of adulteration oils includes vegetable oil, fat oil and margarine at different level were prepared. The extracted properties of the signals obtained from the system were processed using principal component analysis method (PCA) and artificial neural network (ANN). According to the results obtained for the mixture of pure cow ghee with fat oil, vegetable oil, margarine, 97, 96 and 98% of variance by PCA method was obtained. Also, for the ANN method, the ltr">  

  AbstractNowadays, solar energy is one of the cheapest and available renewable energy sources. Among different uses of solar energy, solar collector is one of the most economical ways to use solar energy. In the present study, the outlet water temperature of the solar flat plate collector was modeled using artificial neural networks (ANN) and support vector regression (SVR) and compared with experimental data. Data was collected for 18 days. Water and Bohemite Nano- fluid (ALOOH) with a concentration of 0.2% by weight were used as operating fluid. In order to evaluation of the models, tow structures were tested for both artificial neural network and support vector regression. In the first structure the parameters were input flow, test time, environment temperature and inlet fluid temperature. While in the second structure inputs were input flow, test time, environment temperature, inlet fluid temperature, tow temperatures of absorber plate, and the glass cover temperature. Based on the results, coefficient of determination (R2) and root mean square error (RMSE) in the SVR method for pure water and the first structure, respectively were 0.978991 and 3.2508, respectively, and for the second structure, were 0.998715 and 0.1016, respectively. According to the results, R2 and RMSE for Boehmite Nano fluid and first structure were, 0.958303 and 6/68580, respectively. While these values for the second structure was equal to 0.965097 and 5.4765, respectively. Also, by influencing the type of input fluid as the input of the models, R2 and RMSE were 0.636978 and 281.8210, respectively, and for second structure were 0.939306 and 15.7420. Based on the result for modeling by artificial neural network and for pure water, R2 and RMSE for the first model were 0.99983 and 0.029084, respectively and for the second model were 0.99991 and 0.015617.0, respectively. Also, these values ??for the Bohemite Nano-fluid for the first structure were equal to 0.999 and 0.99896 and for the second structure were 0.9993 and 0.99927 respectively. With the effect of the type of operator fluid as input variable, the R2 and RMSE for the first structure ??were 0.99886 and 0.32567, respectively. Also, these values were 0.99934 and 0.32567, respectively, for the second structure.   Results indicated that the ANN model was better than SVR model for prediction of outlet temperature.   Also by increasing the input parameters, the accuracy of models was increased.Keywords: Artificial neural network, Nano fluid, Outlet fluid temperature, Solar flat plate collector, Support vector regression

AbstractEnergy as one of the most important economic and enviromental factors of production has significant effects and is consididered as a main pillars of development of contries. By considering of coming to the end of fossil fuel, the all countries in the world has an attention to biofuels among all biofuel, has been focused on biodiesel. Production and quality of this fuel is very important issue. Estimination of the trans estrification reaction efficiency is a very important issue in the biodiesel researchs. Therefore in this study evaluated the image processing method and used from response surface methodology (RSM), adaptive neuro-fuzzy inference system (ANFIS) And artificial Neural Network (ANN) to estimination reaction yield. In this study, first biodiesel was produced under different conditions (reaction time, alcohol type (ethanol and methanol), catalyst type (KOH and NaOH)and production method) and reaction yield was achived in the wide   range. Then by using the three methods: 1- Microscopic images 2- Using a special image processing box before centrifugation 3- Using a special image processing box aftre centrifugation, photos was taken from biodiesel sampel in similar conditions. Then by using the image processing provided color channel images in 5 different modes microscopic photos, special compartment photos before centrifugation, combination all methods before centrifugation, special compartment photos after centrifugation   and combination all methods of taking images. By using three modeling methods ANN, RSM and ANFIS, was estimated reaction efficiency properly. The results showed the best way, using of the special image processing compartment without centrifugation by ANFIS modeling method and this method was able to accurately evaluate the trans estrification reaction efficiency with R- squared 0.983, mean squared error 0.002226, mean absolute error 0.02927 and sum squared error 0.12466 performe the best performance in the estimation reaction efficiency. The use of ANN method was able to accurately estimate the reaction efficiency. The use of image processing and ANFIS modeling method, by reducing the cost of analysis compared to the conventional method gas chromatography (GC), estimated reaction efficiency properly.Key words: Biodiesel, Image processing, Gas chromatography (GC), ANN method, Trans esterification, RSM method, ANFIS method.  

  Honey is a natural and sweet substance, which honeybees collect mainly from nectar flowers and process and storage in hives. Adulteration, especially industrial, is made by adding natural syrup or yeast directly to acids. Artificial honey is also made by mixing one or more types of sugar with acid. In this research, the accomplishment construction of a fluid-optimized imaging kit at the Agricultural faculty, Razi University, Kermanshah and fennel honey was also bought from bee keepers in Cangavar. After confirming the origin of honey, 39 samples of adulteration honey using sucrose syrup, fructose syrup and 0.9% fructose syrup with   mixed glucose percentage i   the honey   at 2.5, 5, 7.5, 10, 20, 30, 40, 50, 60, 70, 80, 90 and 100% .   then a natural honey sample, and an artificial honey sample were investigated and compared. In this study, three photographic methods including processing water-soluble honey images (DiW), honey imaging in a special box (black) using petri dish (PD) and microscopic imaging (M), and another adulteration detection method based on physicochemical properties (pH, TDS, EC and MC) and a combination of superior parameters of all mentioned methods was performed. The microscopic and TDS method was distinct from honey type. The (standard ?3). In combination method (C), by using the input parameters of the best model in all previous methods and performing sensitivity analysis, two parameters of (DiW) dissolution method and one parameter of microscopic imaging method (M) were selected and modeled using AFNIS, ANN, and RSM classification systems for hybridization and using the desirability function. The determination coefficient of RSM model was considered 0.9992. Among the best models in all five methods of this research, the RSM model was introduced in the combined method (C) with the least amount of statistical errors compared to other models with the most effective 0.9940 desirability function

Emissions in motor vehicles, especially vehicles with spark ignition engines are the most important environmental parameters of air pollution in many countries. Given the global approach to reducing energy consumption and conservation of energy resources, especially fossil fuels in different countries, restrictions on vehicle fuel consumption has been defined at the international level. Reduction the exhaust gas temperature can be done by fuel enrichment which it will increase the fuel consumption. this problem can be solevd by cooled exhaust gas recirculation. This process will reduce pressure and temperature in the cylinder area during combustion, therefore the tendency to knock   will be reducted. In this study, using the ANSYS 13 software ( CFD section) a model of cooling radiator was analaysid and then based on the results form the model a recirculation circuit was built. The performance of engine temperature and produced emissions in five treatments includes without recirculation, recirculation without cooling and recirculation with cooling with three temperatures (15.4, 11.5 and 7.5 degrees) were test and evaluated. The results of variance analysis indicated a significant difference   for CO, CO2, HC, NOx, and inlet and outlet temperatures by changing the engine speed and recirculation type. Based on the results, the impact of changing the engine speed in the NOx emissions for all modes of recirculation was significant. Also, with increasing the engine speed,   the values of CO and HC emissions, and inlet and outlet air were increased. While with increaseing engine speed, the value of CO2 decreased.