Search

In this study, phenol production process was simulated. Further, the performance of distillation column was optimized through maximizing the mole fraction of cumene in upstream flow. Response surface methodology was applied for design of experiment, modelling, and optimizing the cumene mole fraction in upstream flow of separation column. The analysis of variance was performed for finding the important operative parameters as well as their effects. In this experiment, the effects of three parameters on separation performance were investigated, including number of tray (A), column temperature (B), and reflux ratio (C). Further, radial basis function (RBF) was applied to model the separation column. To develop the neural network model, leave-one-out method was used. This robust model was used for optimizing the performance of separation column. The statistical and artificial intelligence system were capable of predicting mole fraction in upstream flow of distillation column in different conditions with R2 of 0.99 and 0.93, respectively. According to statistical and RBF models, the optimized values of cumene mole fraction are 0.45 and 0.44, respectively.

In this study, a multi-objective design optimization method based on a radial basis function (RBF) model was applied to minimize the volumetric shrinkage and warpage of hip liners as an injection-molded biomedical part. The hip liners included an ultrahigh molecular weight polyethylene (UHMWPE) liner and UHMWPE reinforced with a nano-hydroxyapatite (nHA) liner. The shrinkage and warpage values of the hip liners were generated by simulation of the injection molding process using Autodesk Moldflow. The RBF model was used to build an approximate function relationship between the objectives and the process parameters. The process parameters, including mold temperature, melt temperature, injection time, packing time, packing pressure, coolant temperature, and type of liner, were surveyed to find the interaction effects of them on the shrinkage and warpage of the liners. The results indicated that the addition of nHA helps the liners to obtain more dimensional stability. The model was validated by the k-fold cross validation technique. Finally, the model revealed the optimal process conditions to achieve the minimized shrinkage and warpage simultaneously for various weights.

In recent years, expanding the seafood production has led to dramatic increases in the volume and pollution of wastewater produced by this industry. In this study, anaerobic treatment of seafood processing wastewater was investigated. Further, the performance of treatment process was optimized through maximizing the chemical oxygen demand (COD) removal (%). Several operative parameters such as inlet flow rate, hydraulic retention time (HRT), inlet COD, organic loading rate etc. affect the performance of treatment system. In this experiment, the effects of two parameters on treatment performance were investigated, including HRT and inlet COD. Further, multi-layer feed forward neural network (MLFFNN) and radial basis function (RBF) were applied to model the treatment process. To develop the neural network model, leave-one-out method was used. This robust model was used for optimization the performance of treatment process. The artificial intelligence system were capable of predicting COD removal (%) in different conditions. According to the optimum model, the optimized value of COD removal (%) is 57%.

To make potential the biodegradable polyethylene glycol (PEGO) toward photocrosslinking, it was cooligomerized with fumaryl chloride by esterification reaction. The obtained viscous liquid injectable macromere, OPEGF, was characterized by FTIR and GPC. In another reaction, nanosilica was treated with acryloyl chloride to obtain silica-AC nanopowder and subsequently dispersed in the OPEGO matrix in 1 and 2 wt.% in the presence of either 1,4-butanediol dimethacrylate (BDM) or 1,12-dodecanediol dimethacrylate (DDM) as reactive diluent. The obtained slurries were photocured by blue light irradiation and the characteristics of prepared nanocomposites as well as crosslinked OPEGF were studied via XRD, TEM, tensile, equilibrium swelling and cell culture tests. XRD patterns revealed amorphous nature of all samples. TEM and Si mapping pictures demonstrated good dispersion of silica-AC at nano level (particle size of 20–30nm) in the fabricated samples. Tensile test disclosed a noticeable improvement in mechanical properties of the nanocomposites by increasing in nanofiller level. Equilibrium swelling study revealed that the presence of nanosilica and reactive diluent improved the crosslinking phenomenon. Finally, cell culture test evidenced higher biocompatibility of the samples containing silica-AC nanofiller and BDM reactive diluent. As the consequence of employed tests, designed nanocomposites can be considered as capable candidates for biomedical applications.

The fouling control mechanisms were elucidated in the membrane bioreactor (MBR) by investigating the cake layer specifications in the different hydraulic retention times (HRTs). In this study, petrochemical wastewater was used. The sludge particle size distribution (PSD), excitation-emission matrix (EEM) fluorescence spectra, compressibility cake layer, Fourier transform infrared spectroscopy (FTIR) profile and extracellular polymeric substance (EPS) were measured to determine cake layer characteristics. The results showed that the particle size in the cake layer decreased with reduction in HRT while EPS concentration and transmembrane pressure (TMP) slope increased by time. The EEM florescence spectra of the cake layer showed the existence of two obvious protein-like substance peaks at the wavelength of Ex/Em of 290/355 and Ex/Em of 230–240/355 nm at different HRTs. Furthermore, a feed forward artificial neural network (ANN) was trained using back propagation algorithms for prediction effluent chemical oxygen demand (COD) and TMP. The best structure was a trainlm network with two layers including 17 and 2 neurons in the hidden layer and output layer, respectively. Sensitivity analysis showed that the most and the least sensitive parameters on TMP were mixed liquor suspended solid (MLSS) and time, respectively.

Herein, artificial neural network and adaptive neuro-fuzzy inference system are employed for modeling the effects of important parameters on heat transfer and fluid flow characteristics of a car radiator and followed by comparing with those of the experimental results for testing data. To this end, two novel nanofluids (water/ethylene glycol-based graphene and nitrogen-doped graphene nanofluids) were experimentally synthesized. Then, Nusselt number was modeled with respect to the variation of inlet temperature, Reynolds number, Prandtl number and concentration, which were defined as the input (design) variables. To reach reliable results, we divided these data into train and test sections to accomplish modeling. Artificial networks were instructed by a major part of experimental data. The other part of primary data which had been considered for testing the appropriateness of the models was entered into artificial network models. Finally, predictad results were compared to the experimental data to evaluate validity. Confronted with high-level of validity confirmed that the proposed modeling procedure by BPNN with one hidden layer and five neurons is efficient and it can be expanded for all water/ethylene glycol-based carbon nanostructures nanofluids. Finally, we expanded our data collection from model and could present a fundamental correlation for calculating Nusselt number of the water/ethylene glycol-based nanofluids including graphene or nitrogen-doped graphene.

In this study, a hydrophilic polyethersulfone membrane was used to modify the expensive and low efficient conventional treatment method of wheat starch production that would result in a cleaner starch production process. To achieve a cleaner production, the efficiency of starch production was enhanced and the organic loading rate of wastewater that was discharged into treatment system was decreased, simultaneously. To investigate the membrane performance, the dependency of rejection factor and permeate flux on operative parameters such as temperature, flow rate, concentration, and pH of feed were studied. Response surface methodology (RSM) has been applied to arrange the experimental layout which reduced the number of experiments and also the interactions between the parameters were considered. The maximum achieved rejection factor and permeate flux were 97.5% and 2.42 L min−1 m−2, respectively. Furthermore, a fuzzy inference system was selected to model the non-linear relations between input and output variable which cannot easily explained by physical models. The best agreement between the experimental and predicted data for permeate flux was denoted by correlation coefficient index (R 2) of 0.9752 and mean square error (MSE) of 0.0072 where defuzzification operator was center of rotation (centroid). Similarly, the maximum R 2 for rejection factor was 0.9711 where the defuzzification operator was mean of maxima (mom).

Synthetic wastewater containing wheat starch was treated using sequencing batch reactor (SBR) system working with bio sludge extracted from cow dung. In the first stage, full factorial design was used to investigate the impact of four hydrodynamic variables on process responses and optimum condition was detected. In the second stage, the most important hydrodynamic factor together with four different factors comprised a new design based on the response surface methodology and central composition design. Removal efficiency of starch (COD removal), effluent, and sludge quality were assessed as process responses. Impeller diameter and reactor geometry did not show strong impact on responses. Both designs showed superior results in predicting SBR performance. The analyses revealed that interactions between variables are crucially important and should be accounted in wastewater management. Moreover, it was attained that independent factors have diverse effects on responses which intensifies the necessi...

Various composites including (1-x-y) LiNi0.7Co0.3O2, xLi2MnO3 and yLiCoO2 systems were synthesised using the sol-gel method. Stoichiometric weights of the LiNO3, Mn(Ac)2·4H2O, Co(Ac)2·4H2O, Ni(NO3)2·6H2O for preparing of 28 samples have been used. We found that the sample "Li1.167Ni0.117Co0.699Al0.017Mn0.167O2" with Al doped is the best composition for cathode material in LIBTs. Obviously, the weight of cobalt used in this sample is lower compared to LiCoO2 which is an advantage in view point of cost and toxicity. Charge-discharge characteristics of the mentioned cathode materials were investigated by performing cycle tests in the range of 2.4-4.6 V. Our results confirmed that this kind of system can help removing the mentioned disadvantage of cobalt with high efficiency.

In this study, injection molding process of ultrahigh molecular weight polyethylene (UHMWPE) reinforced with nano-hydroxyapatite (nHA) was simulated and optimized through minimizing the shrinkage and warpage of the hip liners as an essential part of a hip prosthesis. Fractional factorial design (FFD) was applied to the design of the experiment, modeling, and optimizing the shrinkage and warpage of UHMWPE/nHA composite liners. The Analysis of variance (ANOVA) was applied to find the importance of operative parameters and their effects. In this experiment, seven input parameters were surveyed, including mold temperature (A), melt temperature (B), injection time (C), packing time (D), packing pressure (E), coolant temperature (F), and type of liner (G). Two models were capable of predicting warpage and volumetric shrinkage (%) in different conditions with R2 of 0.9949 and 0.9989, respectively. According to the models, the optimized values of warpage and volumetric shrinkage are 0.287222 mm and 13.6613%, respectively. Meanwhile, a finite element analysis (FE analysis) was also carried out to examine the stress distribution in liners under the force values of demanding and daily activities. The Von-Mises stress distribution showed that both of the liners can be applied to all activities with no failure. However, UHMWPE/nHA liner is more resistant to the highest loads than UHMWPE liner due to the effect of nHA in the nanocomposite. Finally, according to the results of injection molding simulations, optimization, structural analysis as well as the tensile strength and wear resistance, UHMWPE/nHA liner is recommended for the production of a hip prosthesis.

In this study the ability of artificial neural network (ANN) in forecasting the daily NASDAQ stock exchange rate was investigated. Several feed forward ANNs that were trained by the back propagation algorithm have been assessed. The methodology used in this study considered the short-term historical stock prices as well as the day of week as inputs. Daily stock exchange rates of NASDAQ from January 28, 2015 to 18 June, 2015 are used to develop a robust model. First 70 days (January 28 to March 7) are selected as training dataset and the last 29 days are used for testing the model prediction ability. Networks for NASDAQ index prediction for two type of input dataset (four prior days and nine prior days) were developed and validated.En este estudio se investigó la capacidad de previsión del índice bursátil diario NASDAQ, por parte de la red neuronal artificial (RNA). Se evaluaron diversas RNA proalimentadas, que fueron entrenadas mediante un algoritmo de retropropagación. La metodología utilizada en este estudio consideró como inputs los precios bursátiles históricos a corto plazo, así como el día de la semana. Se utilizaron los índices bursátiles diarios de NASDAQ del 28 de enero al 18 de junio de 2015, para desarrollar un modelo robusto. Se seleccionaron los primeros 70 días (del 28 de enero al 7 de marzo) como conjuntos de datos de entrenamiento, y los últimos 29 días para probar la capacidad del modelo de predicción. Se desarrollaron y validaron redes para la predicción del índice NASDAQ, para dos tipos de conjuntos de datos de input (los cuatro y los nueve días previos). En este estudio se investigó la capacidad de previsión del índice bursátil diario NASDAQ, por parte de la red neuronal artificial (RNA). Se evaluaron diversas RNA proalimentadas, que fueron entrenadas mediante un algoritmo de retropropagación. La metodología utilizada en este estudio consideró como inputs los precios bursátiles históricos a corto plazo, así como el día de la semana. Se utilizaron los índices bursátiles diarios de NASDAQ del 28 de enero al 18 de junio de 2015, para desarrollar un modelo robusto.Se seleccionaron los primeros 70 días (del 28 de enero al 7 de marzo) como conjuntos de datos de entrenamiento, y los últimos 29 días para probar la capacidad del modelo de predicción. Se desarrollaron y validaron redes para la predicción del índice NASDAQ, para dos tipos de conjuntos de datos de input (los cuatro y los nueve días previos).

The present work deals with application of 25-2 fractional factorial design (FFD) to evaluate the operating parameters on starch separation from synthetic starchy wastewater using a hydrophilic polyethersulfone membrane with 0.65 μm pore size in a plate and frame handmade membrane module. The analysis of variance (ANOVA) combined with F-test was also used to recognize non-significant terms. The performance of the filtration process was evaluated by calculating the COD removal percentage (rejection factor) and permeate flux. In this experiment, five input parameters were surveyed, including trans membrane pressure (TMP), flow and temperature of feed, pH and concentration of wastewater. In this experiment, real wastewater was not used but synthetic starchy wastewater was prepared using starch. Two models were obtained from experimental data, capable of predicting COD removal percentage and permeate flux in different conditions. The predicted values obtained from the regression models were close to the actual ones. For the reduction of fouling, cleaning in place (CIP) method was used.