Tangent Sigmoid Activation Function Research Articles

Predictive modeling of membrane reactor efficiency using advanced artificial neural networks for green hydrogen production

The imperative to decarbonize the energy sector has prompted substantial advancements in clean electricity generation, with hydrogen emerging as a promising low-carbon energy carrier. While hydrogen synthesis from renewable sources is crucial, challenges persist, necessitating innovative approaches for efficient and sustainable production. This study leverages diverse artificial neural network (ANN) models to assess and predict system efficiency based on key operational variables in membrane reactor systems. The multilayered perceptron (MLP) and radial basis function (RBF) methodologies are employed, with the MLP models optimized across twelve training algorithms and eight activation functions, exploring up to three hidden layers with variable neuron counts. The MLP model, utilizing the Levenberg-Marquard training algorithm and Tangent-Sigmoid activation function, achieved a high correlation coefficient (R2) of 0.9975 for training and 0.9962 for testing, and a mean squared error (MSE) of 0.00425 for training and 0.23951 for testing, indicating precise and accurate efficiency predictions. The Log-Sigmoid activation function also performed well, with R² values of 0.9971 (training) and 0.9961 (testing), and MSE values of 0.004086 (training) and 0.17694 (testing). Optimization of the RBF network identified the best performance with a spread parameter of 1 and 35 neurons, although the MLP model demonstrated superior accuracy and reduced computational time. Statistical analysis, encompassing correlation coefficient, mean squared error, Root Mean Squared error, absolute average deviation, absolute average relative deviation, and runtime, confirms the network’s consistent and accurate estimation of system efficiency across various input variables. The study highlights that applying tansig and logsig activation functions, configured with neuron counts of 20, 17, 6 and 23, 20, 2 at the first, second and third hidden layers, respectively, offers enhanced accuracy and reliability. The MLP model’s high performance underscores its potential to identify optimal conditions for H2 generation based on system efficiency, thereby advancing membrane reactor technology for hydrogen production.

Machine learning assisted coupled frequency analysis of skewed multi-phase magnetoelectric composite plates with temperature and moisture dependent properties

In this article, the application of an artificial neural network (ANN)-based machine learning (ML) strategy to predict the coupled frequency of geometrically skewed multiphase magnetoelectric (MME) composite plate exposed to hygrothermal environment is presented. The ANN model is trained using a dataset comprising more than one million simulations conducted using an in-house developed finite element formulation. The underlying multiphysics governing equations are derived using Hamilton’s principle and higher-order shear deformation theory (HSDT). The influence of the hygrothermal environment on the elastic stiffness of MME composites is defined by the empirical constants in the constitutive relations. Four prominent combinations of the empirical constants leading to different elastic stiffness relations have been considered in this study. Alongside, the influence of geometrical skewness on the coupled fundamental frequency is also assessed. For the training of the ANN model, the Levenberg–Marquardt optimization algorithm with 30 neurons along with a tangent sigmoid activation function is used. The trained ANN model is tested on an unseen dataset, different from the training data, and it is shown to accurately predict the natural frequency of MME plate with a maximum error of 2.3%. Further, excluding the training time and considering the computational time alone, the ANN model is found to be 6.3 times faster than the FE simulations. It is anticipated that such ML-based reduced order models can be effective in the design process, especially in complex multiphysics problems, such as the one considered in the work, involving a multitude of geometric, loading and material parameters.

Dry-sliding wear properties of 3D printed PETG/SCF/OMMT nanocomposites: Experimentation and model predictions using artificial neural network

In this article, an attempt has been made to experimentally investigate the synergistic effect of organically modified montmorillonite (OMMT) nanoclay and short carbon fibers (SCFs) on the tribological behaviour of additively manufactured Polyethylene Terephthalate Glycol (PETG) based nanocomposites. The tribo-specimens are 3D printed using fused deposition modelling (FDM). The tribological properties, that is, specific wear rate (SWR) and coefficient of friction (CoF) of various PETG/SCF/OMMT nanocomposites were assessed by performing dry-sliding wear test. In addition, an artificial neural network (ANN) methodology is proposed to accurately predict the wear performance of PETG nanocomposites. The ANN model is trained using the datasets obtained from the experimentation. For the training of the ANN model, the Levenberg–Marquardt optimisation algorithm with 10 neurons along with a tangent sigmoid activation function is utilised. Additional experimentation was performed for arbitrary load and sliding velocity which were not used for training the ANN model and the results were compared to assess the predictive capability of the ANN model on unseen data. The proposed ANN methodology predicted the SWR and CoF with agreeable accuracy. It is believed that by adopting the proposed ANN methodology, the experimentation costs and time can be significantly reduced without compromising on the accuracy of the results.

Evaluating energy absorption of sustainable rubber crumb/kenaf composites through artificial neural network strategies for low‐velocity impact loads

AbstractThis study presents an innovative approach utilizing Artificial Neural Network (ANN) strategies to evaluate the energy absorption capabilities of eco‐friendly rubber crumb/kenaf composites subjected to low‐velocity impact loads. The primary objectives of this research were to assess the energy absorption characteristics of these sustainable composites, understand their mechanical behavior under the impact, and provide valuable insights into their potential applications. To achieve these objectives, an experimental methodology was employed. Rubber crumb/kenaf composites with varying compositions were prepared, and low‐velocity impact tests were conducted using a drop‐weight impact testing apparatus to assess their energy absorption behavior and these parameters were used as inputs for training the ANN models. The weight percentage of waste tire rubber particle (WTRP), type of impactor and impact energies are considered as input data, whereas the absorbed energy is treated as the output. Through the advanced ANN strategies, accurate predictions of energy absorption performance were achieved for the rubber crumb/kenaf composites. The Levenberg–Marquardt optimisation algorithm with ten neurons and a tangent sigmoid activation function is used to train the ANN model. The trained ANN model is tested on an unseen dataset, different from the training data. It is shown to accurately predict the energy absorption characteristics of WTRP/KRE composites with a maximum error of 4.54%. The results revealed that the composite's energy absorption capabilities were influenced by the ratio of rubber crumb to kenaf, as well as the impact velocity. Additionally, the ANN models demonstrated excellent predictive capabilities, enabling efficient estimation of energy absorption behavior. The significance of these results lies in the potential applications of eco‐friendly rubber crumb/kenaf composites. By understanding their energy absorption characteristics, these composites can be effectively utilized in various industries. For instance, they could be employed in automotive parts manufacturing to enhance occupant safety during low‐velocity impact events. Furthermore, these composites can find applications in sports equipment, protective gear, and other impact‐prone products, contributing to sustainable and environmentally friendly materials. It is believed that by adopting the proposed ANN methodology, the experimentation costs and time can be significantly reduced without compromising the accuracy of the results. The obtained results provide valuable insights into the mechanical behavior of these sustainable composites and open avenues for their implementation in diverse industries where impact resistance is crucialHighlights Development of sustainable Rubber crumb/kenaf composites Application of advanced ANN based predictive assessment of the impact response Parametric study of impactor shape and height of impact is performed The damage mechanism contributing to energy absorption is micrographically studied

Artificial intelligence techniques for encrypt images based on the chaotic system implemented on field-programmable gate array

<span lang="EN-US">Image encryption is an important issue in protecting the content of images and in the area of information security. This article proposes a novel method for image encryption and decryption using the structure of the artificial neural network (ANN)-based chua chaotic system (CCS). This structure was efficiently designed on a field-programmable gate array (FPGA) chip utilizing the xilinx system generator (XSG) tool with the IEEE-754-1985 32-bit floating-point number format. For ANN-based CCS design, a multilayer feed forward neural network (FFNN) structure with three inputs and three outputs was created. This structure consists of one hidden layer with four neurons, each of which has a Tangent Sigmoid activation function. The training of ANN-based CCS yielded a 3.602e-13 mean square error (MSE) value. After successfully training the ANN-based CCS, the design was carried out on FPGA, utilizing the ANN structure's bias and weight values as a reference. The xilinx vivado (2017.4) design suite was used to synthesis and test the ANN-based CCS on the FPGA. The histogram, correlation coefficient, and entropy are used to perform security analysis on various images. Finally, FPGA hardware co-simulation using a Xilinx Artix7 xc7a100t-1csg324 chip was utilized to verify that the encryption and decryption of the images were successful.</span>

Modeling of stage-discharge using back propagation ANN-, ANFIS-, and WANN-based computing techniques

The development of the stage-discharge relationship is a fundamental issue in hydrological modeling. Due to the complexity of the stage-discharge relationship, discharge prediction plays an essential role in planning and water resource management. The present study was conducted to model discharge at the Gaula Barrage site in the Uttarakhand state of India. The study evaluated adaptive neuro-fuzzy inference system (ANFIS)-, artificial neural network (ANN)-, and wavelet-based artificial neural system (WANN)-based models to estimate the discharge. The daily data of 12 years (2007–2018) were used to train and test the models. The gamma test was used to identify the best model for discharge prediction. The input data having a stage with 1-day lag and discharge with 1- and 2-day lags and current-day discharge as output was used for discharge modeling. In the case of ANN models, the back propagation algorithm and hyperbolic tangent sigmoid activation function were used. WANN used Haar, a trous-based wavelet function. In ANFIS models, triangular, psig, generalized bell, and Gaussian membership functions were used to train and test the models. The models were evaluated qualitatively and quantitatively using correlation coefficient, and root means square error, Willmott index, and efficiency coefficient. It was found that the ANFIS model performed better than ANN- and WANN-based models for discharge prediction at the Gaula Barrage.

Multi-Objective Optimization Through Machine Learning Modeling for Production of Xylooligosaccharides from Alkali-Pretreated Corn-Cob Xylan Via Enzymatic Hydrolysis.

The hemicellulose content present in corn cobs can help in producing a high amount of xylooligosaccharides (XOS) in an eco-friendly manner. In this work, the XOS was produced from alkali pre-treated corn-cobs having a true yield of 38 ± 1.4% via enzymatic hydrolysis with the help of xylanase from T. lanuginosus VAPS-24. The production process was optimized to achieve a high concentration of XOS using innovative multi-objective optimization through machine learning modeling and finding out the most suitable parameters where xylobiose production is higher than xylose. The Multi-objective connected neural networks (MOCNN) model with tangent sigmoid activation function yielded a correlation coefficient of 96.51%; there were six optimal sets where xylobiose concentration was higher than xylose. The best-optimized conditions yielded 3.03mg/ml of xylobiose and 1.31mg/ml of xylose. Therefore, this novel approach of machine learning can target the increasing demand for xylooligosaccharides in the growing industrial market of prebiotics.

Modeling the energy gain reduction due to shadow in flat-plate solar collectors; Application of artificial intelligence

Energy lost due to shadow in the absorber plate of solar collectors can decrease the solar energy gain. In some studies, mathematical modeling was applied for calculating the energy gain reduction due to shadow in flat-plate solar collectors. In this study, ANN method was developed for modeling the energy gain reduction. Multilayer Perceptron (MLP) with one hidden layer and a range of neurons (5–30) by two training algorithms (LM and BR) and tangent sigmoid activation function (TanSig) were used by help of K-fold cross validation method. In the first section, six set of solar collector dimensions were used (1×1; 1×1.5; 1×2; 1.5×1.5; 1.5×2 and 2×2). In the second section all the range of dimensions were used as the inputs. The results of the first section showed that MLP with BR training algorithm can predict the energy gain reduction with minimum MAPE and RMSE in all the categories. The best results related to (1.5×1.5) dimension that achieved a MAPE of 0.15 ± 0.09% and RMASE of 4.42 ± 2.43 KJm−2 year−1, respectively. The results of the second section indicated that BR is a better training algorithm than LM. The MAPE and R2 factors for the best topology (5-27-1) were 0.0610 ± 0.0051% and 0.9999 ± 0.0001, respectively. The results of the sensitivity analysis showed that height has the biggest impact on total energy gain reduction due to shadow in flat-plate solar collectors. Finally, the results of this study indicated that by using ANN and decrease the energy lost, the efficiency of solar collectors can be increased in all applications such as industry and agriculture.

Optimal artificial neural network design for simultaneous modeling of multicomponent adsorption

In this work, ultrasound-modified chitin (UM-chitin) was used to adsorb cobalt (Co2+), methylene blue (MB), and nickel (Ni2+) in single, binary and ternary systems, at different temperatures. The single adsorption isotherms showed that the affinity of the UM-chitin towards the metals was higher than that of the dye. Antagonist adsorption was found for the binary adsorption of Co2+ and MB; however, the presence of Ni2+ in the ternary system promoted the adsorption of MB and inhibited the adsorption of Co2+. Therefore, an optimal artificial neural network was developed to predict these synergistic and antagonistic interactions simultaneously. Different ANN configurations were tested using the initial concentration of each adsorbate (0–650 mg L−1) and the solution temperature (298–328 K) as input variables. The optimal ANN structure, which uses 2 hidden layers with 5 and 10 neurons, tangent sigmoid activation function (tansig) at both hidden and output layers, and Bayesian Regulation as its backpropagation algorithm demonstrated suitable ability in simultaneously predict the adsorption capacity for each adsorbate (MSE < 0.0003 and R > 0.9995).

Relating groundwater levels with meteorological parameters using ANN technique

In this study, the groundwater occupied within the boundaries of River Ravi and River Sutlej was investigated by using artificial neural networks (ANN) by using the meteorological parameters (data of rainfall (R), maximum temperature (Max.T), minimum temperature (Min.T), solar radiation (S.R.), wind (W), relative humidity (R.H.), the elevation of area (E), polygon area (A) and water table depths (D/W). The best efficient models were studied by using different types of network architecture, such as the number of neurons, hidden layers, and activation function with a different percent of data in training, validation, and testing. The developed Levenberg-Marquardt back-propagation ANN models were compared through statistical performance criteria: Mean Square Error (MSE), Mean Absolute Error (MAE) and Coefficient of determination (R). The results (TT-8-24-1 for pre-monsoon and TT-8-24-1 for post-monsoon) show that ANN model with single hidden layer, 24 neurons, 80% of data for training, 10% for validation, 10% for testing and using tangent sigmoid activation function was found to be optimistic ANN model with MAE, MSE and R values of 0.0338, 0.0023 and 0.97 for pre-monsoon and 0.031, 0.0021 and 0.974 for the case of post-monsoon respectively.

Artificial Neural Network-Based 4-D Hyper-Chaotic System on Field Programmable Gate Array

In this presented study, a 4-D hyper-chaotic system newly proposed to the literature, has been implemented as Multi-Layer Feed-Forward Artificial Neural Network-based on FPGA chip with 32-bit IEEE-754-1985 floating-point number standard to be utilized in real time chaos-based applications. In the first step of the study, 4-D hyper-chaotic system has been numerically modeled on FPGA using Dormand-Prince numeric algorithm. In the second step, the data set (4X10,000) obtained from Matlab-based numeric model has been divided into two parts as training data set (4X8,000) and test data set (4X2,000) to create ANN-based 4-D hyper-chaotic system. A Multi-Layer Feed-Forward ANN structure with 4 inputs and 4 outputs has been constructed for ANN-based 4-D hyper-chaotic system. This structure has only one hidden layer and there are 8 neurons having Tangent Sigmoid activation function used as the activation function in each neuron. 2.58E-07 Mean Square Error (MSE) value has been obtained from the training of ANN-based 4-D hyper-chaotic system. In the third step, after the successful training of ANN-based 4-D hyper-chaotic system, the design of ANN-based 4-D hyper-chaotic system has been carried out on FPGA by taking the bias and weight values of the ANN structure as reference. In this step, at first, Matlab-based Feed-Forward Multi-Layer 4X8X4 network structure has been coded in Very High Speed Integrated Circuit Hardware Description Language (VHDL) to be implemented on FPGA chips. Then, the bias and weight values of the ANN structure has been converted from decimal number system to floating-point number standard and these converted values have been embedded into the network structure. In the last step, the ANN-based 4-D hyper-chaotic system designed on FPGA has been synthesized and tested using Xilinx ISE Design Suite. The chip statistics have been given after the Place&Route process carried out for the Virtex XC6VHX255T-3FF1155 FPGA chip. The maximum operating frequency of ANN-based 4-D hyper-chaotic system on FPGA has been obtained as 240.861 MHZ.

Shallow Landslide Susceptibility Models Based on Artificial Neural Networks Considering the Factor Selection Method and Various Non-Linear Activation Functions

Landslide susceptibility mapping is well recognized as an essential element in supporting decision-making activities for preventing and mitigating landslide hazards as it provides information regarding locations where landslides are most likely to occur. The main purpose of this study is to produce a landslide susceptibility map of Mt. Umyeon in Korea using an artificial neural network (ANN) involving the factor selection method and various non-linear activation functions. A total of 151 historical landslide events and 20 predisposing factors consisting of Geographic Information System (GIS)-based morphological, hydrological, geological, and land cover datasets were constructed with a resolution of 5 x 5 m. The collected datasets were applied to information gain ratio analysis to confirm the predictive power and multicollinearity diagnosis to ensure the correlation of independence among the landslide predisposing factors. The best 11 predisposing factors that were selected in this study were randomly divided into a 70:30 ratio for training and validation datasets, which were used to produce ANN-based landslide susceptibility models. The ANN model used in this study had a multi-layer perceptron (MLP) structure consisting of an input layer, one hidden layer, and an output layer. In the output layer, the logistic sigmoid function was used to represent the result value within the range of 0 to 1, and six non-linear activation functions were used for the hidden layer. The performance of the landslide susceptibility models was evaluated using the receiver operating characteristic curve, Kappa index, and five statistical indices (sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV)) with the training dataset. In addition, the landslide susceptibility models were validated using the aforementioned measures with the validation dataset and were compared using the Friedman test to check the significant differences among the six developed models. The optimal number of neurons was determined based on the aforementioned performance evaluation and validation results. Overall, the model with the best performance was the MLP model with the logistic sigmoid activation function in the output layer and the hyperbolic tangent sigmoid activation function with five neurons in the hidden layer. The validation results of the best model showed a sensitivity of 82.61%, specificity of 78.26%, accuracy of 80.43%, PPV of 79.17%, NPV of 81.82%, a Kappa index of 0.609, and AUC of 0.879. The results of this study highlight the effectiveness of selecting an optimal MLP model structure for shallow landslide susceptibility mapping using an appropriate predisposing factor section method.

Landfill area estimation based on solid waste collection prediction using ANN model and final waste disposal options

Public health of inhabitants has been affected by the increase of unsound waste management in the cities of developing countries. Solid waste management has received extensive attention from the municipalities. A sustainable plan and design of solid waste management system of a city necessitate accurate forecasting of the solid waste generated and collected by the city authorities in the landfill for final disposal and other renewable energy options. In the study, an Artificial Neural Network (ANN) surrogate model was used to predict solid waste collected from the year 2012–2016 at Matuail landfill site of Dhaka South City Corporation (DSCC). 120 monthly solid waste quantity and vehicle trip number getting from weighbridge were used as inputs data into the model. 70% of the data used for the model training, 15% of the data used for validation and 15% of that used for testing. The remaining 60 monthly waste quantities are used as output to develop the model. Feed-forward back propagation neural network was used with the hyperbolic tangent sigmoid activation function and the Levenberg–Marquardt optimization method. The model with 2-5-1-1 topology is selected as the best topology based on the performance metrics i.e., the minimum value of MSE and high value of regression. The ANN based solid waste forecasting model performed to be promising using the available weighbridge data with the coefficient of determination (R2) for training and testing 0.85 and 0.86. The developed model can be alternatively used successfully with weight bridge software in the landfill to efficiently forecast solid waste collection, particularly that in the countries with a similar demographic and social environment. Considering other proposed alternative disposal options and waste characterization the required landfill area are estimated and found that the landfill authority can save the valuable urban landfill area requirement up to 28.6%. The result shows that the innovative proposed method of landfill area estimation by ANN and final disposal methods can be used alternatively that helps for better planning and management of the landfill site.

Ascertaining of chatter stability using wavelet denoising and artificial neural network

In modern machining industries, tool chatter detection and suppression along with maximized metal removal rate is a challenging task. Inexpedient vibration between cutting tool and work piece promotes unstable cutting. This results in enhanced detritions of tool and poor surface finish along with unpredictable metal removal rate. In the present work, effect of machining parameters such as depth of cut ( d), feed rate ( f) and spindle speed ( N) on chatter severity and metal removal rate have been ascertained experimentally. Experimentally recorded raw chatter signals have been denoised using wavelet transform. An artificial neural network model based on feed forward back propagation network has been proposed for predicting stable cutting zone and metal removal rate in turning process. It has been deduced that Tangent Sigmoid activation function in an artificial neural network is the best option to achieve the aforesaid objectives. Well correlation between the artificial neural network predicted results and experimental ones validate the developed technique of ascertaining the tool chatter severity.

Artificial Intelligence Treatment of SO2 Emissions from CFBC in Air and Oxygen-Enriched Conditions

Because the complexity of sulfur capture and release during solid-fuel combustion in a circulating fluidized bed combustors (CFBC), especially in the oxygen-enriched combustion, has not been sufficiently recognized, the development of a simple model, which can correctly predict the SO2 emissions from such units over a wide range of operating conditions is of practical significance. The artificial neural network (ANN) approach is proposed in this paper, which may overcome the shortcomings of the experimental procedures and the programmed computing approach. The Ca:S molar ratio, oxygen concentration in inlet gas, excess oxygen, average riser temperature, mean diameter of the coal particles, average gas velocity in the riser, flue gas recycle ratio, and inlet gas pressure are taken into account by the model as the input parameters. The [8-3-7-1] ANN model with hyperbolic tangent sigmoid activation function was successfully applied to calculate the SO2 emissions from coal combustion in several CFB boilers operating under both air-fired and oxygen-enriched conditions.

Modeling of Drying Process of Bittim Nuts (pistacia terebinthus) in a Fixed Bed Dryer System by Using Extreme Learning Machine

Abstract Artificial neural networks (ANNs) have been widely used in modeling of various systems. Training of ANNs is commonly performed by backpropagation based on a gradient-based learning rule. However, it is well-known that such learning rule has several shortcomings such as slow convergence and training failures. This paper proposes a modeling technique based on Extreme Learning Machine (ELM) eliminating disadvantages of backpropagation based on a gradient-based learning rule for the drying of bittim (pistacia terebinthus). The samples for ELM based model are obtained by experimental studies. In experimental studies, the sample mass loss rate as a function time was investigated in different air velocities (0.5 and 1 m/s) and air temperatures (40, 60 and 80°C) in a designed dryer system. The obtained samples from experiments are used for training and testing of ELM. Further, some parameters of ELM such as type of activation function and the number of hidden neurons are set to obtain the best possible modelling results. The obtained prediction results show that ELM algorithm with tangent sigmoid activation function and 20 hidden neurons is appeared to be most optimal topology since maximum R2 and minimum rms (0.0500) and cov (0.2256) values are obtained. Thus, it is concluded that ELM can be used as an effective modelling tool in the drying of bittim (pistacia terebinthus) in fixed bed dryer system.

MLP-based drought forecasting in different climatic regions

Water resources management is a complex task and is further compounded by droughts. This study applies a multilayer perceptron network optimized using Levenberg–Marquardt (MLP) training algorithm with a tangent sigmoid activation function to forecast quantitative values of standardized precipitation index (SPI) of drought at five synoptic stations in Iran. The study stations are located in different climatic regions based on De Martonne aridity index. In this study, running series of total precipitation corresponding to 3, 6, 9, 12, and 24 months were used and the corresponding SPIs were calculated: SPI3, SPI6, SPI9, SPI12, and SPI24. The multilayer perceptrons (MLPs) for SPIs with the 1-month lead time forecasting, were tested and validated. Four different input vectors were considered during network development. In the first model, MLP constructed by importing antecedent SPI with 1-, 2-, 3-, and 4-month time lags and antecedent precipitation with 1- and 2-month time lags (MLP1). Addition of antecedent North Atlantic Oscillation or antecedent Southern Oscillation Index with 1-month time lag or both of them to MLP1 led to MLP2, MLP3, and MLP4, respectively. The MLP models were evaluated using the root mean square error (RMSE) and the coefficient of determination (R2). The results showed that MLP4 had a higher prediction efficiency than the other MLPs. The more satisfactory results of RMSE and R2 values of MLP4 for 1-month lead time for validation phase were equal to 0.35 and 0.92, respectively. Also, results indicated that MLPs can forecast SPI24 and SPI12 more accurately than the other SPIs.

Forecasting Extreme PM10 Concentrations Using Artificial Neural Networks

Life style and life expectancy of inhabitants have been affected by the increase of particulate matter 10 micrometers or less in diameter (PM10) in cities and this is why maximum PM10 concentrations have received extensive attention. An early notice system for PM10 concentrations necessitates an accurate forecasting of the pollutant. In the current study an Artificial Neural Network was used to estimate maximum PM10 concentrations 24-h ahead in Tehran. Meteorological and gaseous pollutants from different air quality monitoring stations and meteorological sites were input into the model. Feed-forward back propagation neural network was applied with the hyperbolic tangent sigmoid activation function and the Levenberg–Marquardt optimization method. Results revealed that forecasting PM10 in all sites appeared to be promising with an index of agreement of up to 0.83. It was also demonstrated that Artificial Neural Networks can prioritize and rank the performance of individual monitoring sites in the air quality monitoring network.

An Artificial Neural Network Model for Design of Wellhead Chokes in Gas Condensate Production Fields

The subcritical flow behavior of gas condensates through wellhead chokes under different flow conditions are studied by use of an artificial neural network (ANN). The proposed network is trained using the Levenberg-Marquardt back-propagation algorithm and the hyperbolic tangent sigmoid activation function is applied to calculate the output values of the neurons of the hidden layer. The proposed neuromorphic model outperforms the existing empirical correlations both in accuracy and generality. The results of this work are very important in the design of wellhead chokes under a wide range of flow conditions usually encountered during the flow of gas condensates.

Accurate prediction of the dew points of acidic combustion gases by using an artificial neural network model

This paper presents a new approach based on using an artificial neural network (ANN) model for predicting the acid dew points of the combustion gases in process and power plants. The most important acidic combustion gases namely, SO 3, SO 2, NO 2, HCl and HBr are considered in this investigation. Proposed Network is trained using the Levenberg–Marquardt back propagation algorithm and the hyperbolic tangent sigmoid activation function is applied to calculate the output values of the neurons of the hidden layer. According to the network’s training, validation and testing results, a three layer neural network with nine neurons in the hidden layer is selected as the best architecture for accurate prediction of the acidic combustion gases dew points over wide ranges of acid and moisture concentrations. The proposed neural network model can have significant application in predicting the condensation temperatures of different acid gases to mitigate the corrosion problems in stacks, pollution control devices and energy recovery systems.