Artificial Neural Network Method Research Articles

Use of machine learning methods in predicting the main components of essential oils: Laurus nobilis L.

Laurus nobilis L. is an export plant that contributes to the industry with its medicinal and aromatic properties, as well as the use of its leaves or fruits in the food, pharmaceutical, and cosmetic industries. This study aims to predict the first three main components of bay laurel essential oil with different machine learning methods as a result of GC-MS analysis, taking into account a wide range of parameters used before and after the analysis of the essential oil in pharmacognostic studies. In this study, the K-Means algorithm was used to detect unknown relationships (chemical or physical properties) between the data obtained as a result of the analysis, and Artificial Neural Network (ANN) and Decision Tree machine learning methods were used to perform prediction operations. Based on this, considering the R2 values obtained from ANN analyses, R2 values of 0.9992 for testing, 1 for training, 1 for validation value, and 0.99992 for all were obtained. The ANN model produced results close to the real values with an accuracy rate of 98.97% in predicting the three main components. The Decision Tree algorithm achieved classification with a 96.23% accuracy rate in predicting the three main components. K-Means clustering was performed with an accuracy rate of 97.83% and significant relationships between the features of the Laurus nobilis dataset were detected. This comprehensive study serves as a guide for machine learning research on other essential oils.

QSPR-based prediction model for the melting point of polycyclic aromatic hydrocarbons using MLR and ANN methods

The melting point is an important property that helps generate specific compounds with desired thermos-physical properties. Much work has been done applying quantitative structure-property relationships to improve the melting-point correlations, but they are unreliable. This gap might come from the melting point's sensitivity for small molecular variations and descriptors, which currently do not fully consider all factors determining melting behavior. In this work, we provide a QSPR model for predicting the melting point of a heterogeneous polycyclic aromatic hydrocarbons dataset. The model was generated using a robust hybrid linear approach (Genetic Algorithm-Multiple Linear Regression) and a nonlinear approach named Artificial Neural Network (ANN). Three descriptors were chosen to explain the influence of molecular weight and symmetry on melting point. The resulting QSPR model can model melting-point behavior with an RMSE of 34.88K, a coefficient correlation value of R²=0.887, and a prediction coefficient of Q²LOO= 0.863. This study reveals that the results produced by MLR were appropriate and served to predict melting points. However, compared to the results obtained by the ANN model, we conclude that the latter is more effective and better than the MLR model. Based on the results, our suggested model may be effective in predicting melting points, and the selected descriptors play essential roles in determining melting points.

Pore pressure prediction of hydrocarbon reservoirs with empirical models and artificial neural network: case study in the Doba basin, Chad

Accurate Pore pressure prediction is essential in drilling as it enables drilling hazards to be avoided. However, traditional methods of pore pressure prediction, especially empirical methods are field-specific. It is therefore important to evaluate the best empirical method which is to be applied in any basin or field according to its main characteristics. Besides, recently, artificial neural networks (ANNs) are the most used for pore pressure prediction because they can help to learn complex relationships between input and output parameters. They lead to accurate pore pressure prediction results in many cases. In light of this, the aim of this study carried out in the Doba basin is to analyse and compare the applicability of various sonic log-based empirical models of pore pressure prediction, including Eaton’s model, Raymer-Hunt’s model and Kozeny-Carman’s model and a ANN model of pore pressure prediction in order to find out the best predictive one to use in the studied area in order to achieve better prediction results. To achieve this objective, firstly, the mathematical models of pore pressure prediction from Eaton, Terzaghi and Raymer-Hunt equations and Kozeny-Carman’s model were used. The artificial neural network method uses 33 datasets of well log data including sonic transient time, vertical stress and porosity as input parameters to train the ANN model. The output parameters are the measured pore pressures produced by the Repeat Formation Tester. Using 70% of data set for training, 15% for test and 15% for validation, the trained model with the Bayesian Regularization Backpropagation function predicts pore pressure with an average correlation coefficient of 98%. The results show that pore pressure obtained from the Kozeny-Carman’s model, also has a 98% correlation coefficient, thus proving to be the most accurate of empirical methods. Indeed, pore pressures obtained from Raymer-Hunt’s model and Eaton’s model have 97% and 90% respectively as correlation coefficients. This study provides the best and most reliable empirical method of sonic log-based pore pressure prediction to use in the Doba basin, which can be used to achieve the best pore pressure prediction results and hence ensures drilling safety and wellbore stability in the studied area. The study also highlights a good application of the Bayesian function of the ANN for pore pressure prediction in the investigated area.

Intelligent frequency control of AC microgrids with communication delay: An online tuning method subject to stabilizing parameters

Smart control techniques have been implemented to address fluctuating power levels within isolated microgrids, mitigating the risk of unstable frequencies and the potential degradation of power supply quality. However, a challenge lies in the fact that employing these computationally complex methods without stability preservation might not suffice to handle the rapid changes of this highly dynamic environment in real-world scenarios over communication delays. This study introduces a flexible real-time approach for the frequency control problem using an artificial neural network (ANN) constrained to stabilized regions. Our solution integrates stabilizing PID controllers, computed through small-signal analysis and tuned via an automated search for optimal ANN weights and reinforcement learning (RL)-based selected constraints. First, we design stabilizing PID controllers by applying the stability boundary locus method and the Mikhailov criterion, specifically addressing communication delays. Next, we refine the controller parameters online through an automated process that identifies optimal coefficient combinations, leveraging a constrained ANN to manage frequency deviations within a restricted parameter range. Our approach is further enhanced by employing the RL technique, which trains the tuning system using an interpolated stability boundary curve to ensure both stability and performance. This one-of-a-kind combination of ANN, RL, and advanced PID tuning methods is a big step forward in how we handle frequency control problems in isolated AC microgrids. The experiments show that our solution outperforms traditional methods due to its reduced parameter search space. In particular, the proposed method reduces transient and steady-state frequency deviations more than semi- and unconstrained methods. The improved metrics and stability analysis show that the method improves system performance and stability under changing conditions.

Machine learning optimization and 4E analysis of a CCHP system integrated into a greenhouse system for carbon dioxide capturing

The world has seen an increase in the popularity of renewable-based energy systems. However, because of their erratic nature, fossil fuels cannot be entirely phased out. Utilizing carbon dioxide in greenhouses close to power plants and generating extra power electricity from dissipated thermal energy are appealing strategies that cause us to have a cleaner environment and affordable power electricity. To exploit the high enthalpy of expelled gases from a gas turbine, the gas turbine was coupled with an absorption refrigerant cycle and an organic Rankine cycle and analyzed. Additionally, a deep artificial neural network method was used to calculate all functions in a rapid optimization and accurate 4E analysis. The carbon dioxide emissions subsided from 0.9183 to 0.41 kg CO2/s, and the optimization time improved 257 times after using the trained machine learning model. An adsorption technique for CO2 separation is used for the proposed system because of its lower energy consumption. The maximum exergy and energy efficiencies were attained at 36.71 % and 49.2 %, respectively, and parametric studies are being assessed. Due to increases in harvests and efficiencies, the net annual interest has increased by 21.8 %, up to 22.5 million dollars.

Digitalization of Analysis of a Concrete Block Layer Using Machine Learning as a Sustainable Approach

The concrete block pavement (CBP) system has a surface layer consisting of concrete block pavers and joint sand over a bedding sand layer. The non-homogeneous nature of the surface course of CBP, along with different laying patterns and shapes of block pavers, makes the analysis of CBP cumbersome. In this study, the surface course of CBP was modeled based on the slab action of the block pavers and joint sand, which are connected together in full contact. Four different laying patterns, including herringbone, stretcher, parquet, and square, were modeled using a finite element model. The elastic moduli of the block pavers varied from 2500 MPa to 45,000 MPa, with thicknesses ranging from 60 mm to 120 mm. As a result, modeling of CBP based on slab action can be considered a realistic strategy. In addition, a dataset was created based on quantitative inputs, e.g., elastic modulus and thickness of the block pavers, and qualitative input, i.e., block laying patterns. The approaches of machine learning adopted were support vector regression, Gaussian process regression, single-layer and deep artificial neural networks, and least squares boosting to implement prediction approach based on input and output. The analyses of statistical accuracy of all five machine learning methods showed high accuracy; however, the Gaussian process and deep artificial neural network methods resulted in the most accurate outputs and are recommended for further studies. Based on the machine learning models, digitalization is achieved through the development of simple, user-friendly software for electronic devices in order to perform a preliminary analysis of different laying patterns of CBP. Such a platform may result in less laboratory work and boosts the level of sustainability in concrete block pavement technology.

Effective Electromagnetic Properties of Composite Material Computed From Neural Network Approach

ABSTRACTThanks to their lightweight, composite materials have become widely used in the automotive and aerospace industries. The design of components made from these materials is carried out by numerical modeling which can sometimes be tedious because of the need to take into account the internal structure of these materials. Obtaining the effective properties of an equivalent homogeneous material to replace the composite in our numerical models makes modeling easier. Classical homogenization approaches are not always suitable to obtain these effective properties. This article deals with an inverse problem that consists in computing the electromagnetic properties from the knowledge of the magnetic shielding effectiveness values. For different composite samples, an artificial neural network method is used to predict the effective conductivities from the magnetic shielding effectiveness measurements. The magnetic shielding effectiveness values computed from the predicted conductivities are close to those obtained from the measurements.

Artificial Neural Network Model For Optimization of Forecasting Material Inventory

The increasing competition in the fast-moving consumer goods (FMCG) industry leads to demand fluctuations, negatively impacting the accuracy of demand forecasts and determining optimal lot sizes in material inventory planning. Many companies struggle to adopt appropriate forecasting models, resulting in poor accuracy and higher material costs. This study aims to develop an integrated model for forecasting and material planning using simulation. The artificial neural network (ANN) method is proposed to improve forecasting accuracy, with performance evaluated through mean percentage error (MAPE), mean absolute deviation (MAD), and mean squared error (MSE). The forecast results are then applied to optimize material inventory using the economic order quantity (EOQ) model, considering warehouse capacity constraints. The EOQ model is applied to adjust lot sizes under time-varying demand. The findings highlight the importance of integrating forecasting with inventory planning to provide accurate demand predictions and optimal lot sizing, ultimately minimizing material costs in the FMCG industry. This research contributes to better decision-making in supply chain management by enhancing forecasting accuracy and inventory optimization.

Classification of different wheat flour types using hyperspectral imaging and machine learning techniques

Different wheat flour types are used to produce various baked products. Due to the whiteness of the four types, hyperspectral imaging can be used due to receiving infrared wavelength. The technique was applied to distinguish confectionery flour and the flours of Samoun, Sangak, and Tafton breads using a line scanning system in the range of 400–950 nm. Effective wavelengths were selected and different image features were extracted from the corresponding image channels. The selected wavelengths were 601.33, 620.34, 696.41, 730.31, 821.26, and 841.11 nm. The extracted features were used in classification step using linear discriminant analysis, support vector machine, and artificial neural network methods in MATLAB software. The classification accuracy of artificial neural network was higher than the other methods. The efficient features gave higher classification accuracy (98.1 %) than all extracted features (96.9 %). The results showed the high ability of hyperspectral imaging combined with artificial neural network to distinguish different wheat flour types.

ONION CRACKERS SALES FORECASTING USING ARTIFICIAL NEURAL NETWORK METHOD AND HOLT'S DOUBLE EXPONENTIAL SMOOTHING

Changes in product demand is a problem that is often faced by the industry as well as one of them is onion crackers. Tapioca flour is the main ingredient used to make onion crackers. Because the demand for crackers is always changing, this company often experiences excess or shortage of raw materials. If there is an excess of raw materials, the company must incur additional costs for the maintenance and storage of raw materials so that raw materials can be properly stored in accordance with existing standards, which of course costs a lot. Therefore, companies must plan to solve this problem by planning raw material requirements by forecasting raw material requirements using the artificial neural network method and double exponential smoothing holt. The results showed that the artificial network method had a mean square error of 0.120 and the mean square error using the double exponential smoothing method yielded a value of 206.19. Based on these two values, it can be concluded that the artificial neural network method is more accurate than the double exponential smoothing holt method. This can be seen by comparing the roat mean square error values of the two methods..

ANN modeling of tincal ore dehydration

Abstract Tincal ore is a preferred material in many industrial applications, especially without water. It is important to dehydrate boron ores so that they can be used in materials engineering. For this purpose, some heat treatments must be carried out. Heat treatments are associated with additional costs. It is possible to model heat treatments using artificial intelligence methods, determine optimal process conditions and achieve the desired results with much less processing effort. In this study, a dehydration process was first carried out to dehydrate tincal ore and ANN (artificial neural networks) modeling of this process was investigated using the parameters of temperature, time and amount of ore. The possibility of achieving the desired H2O, B2O3 and Na2O concentration values in the dewatering process in the shortest time and by the shortest route was investigated using the ANN model. In the modeling, a single model was designed for the changes in concentrations and this model was trained separately for each. The result of the modeling was that the R 2 values for all three models were close to each other and were approximately 0.98. It was thus shown that the ANN method can be successfully modeled for dewatering processes.

Interfacial stress analysis in damaged RC beams reinforced whit FRP using artificial neural network method

Interfacial stress analysis in damaged RC beams reinforced whit FRP using artificial neural network method

Analysis of Heartbeat Signals to Detect Sleep Disorders Using Artificial Neural Network Methods

A human sleep disorder detection system has been designed using an AD8232 Electrocardiogram sensor module integrated with a microcontroller and internet connection through ESP 32. The heartbeat signals from the sensor are analyzed using Artificial Neural Network (ANN) methods to determine normal conditions, Obstructive Sleep Apnea (OSA), or Central Sleep Apnea (CSA). The sensor's accuracy was measured over 10 measurements, resulting in 96.85%. Testing with 30 training data samples achieved an accuracy of 93.33%, and testing with 20 training data samples achieved an accuracy of 80%. The system displays output values through the Internet of Things (IoT) with an average computation time of around 7.6 ms.

Prediction Of Student Achievement Using Artificial Neural Network And Support Vector Regression At SMK TELKOM Lampung

The analysis of student performance is crucial in vocational schools because it helps identify the challenges students face in preparing themselves for the workforce. By integrating data mining techniques such as Artificial Neural Networks (ANN), educators can enhance their understanding of factors that improve student learning outcomes. An artificial neural network (ANN) is composed of interconnected artificial neurons that can learn from input data and make complex predictions, including academic achievements. The structure and function of the human brain inspire ANN. This study compares the effec- tiveness of the artificial neural network (ANN) method with other methodologies, such as support vector regression (SVR), to predict student achievement at SMK Telkom Lampung. Primary data collected from SMK Telkom Lampung includes 4939 examples with 550 cases, 26 features, and 4 meta-attributes. Performance evaluation involves metrics such as Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and coefficient of determination (R2). The coefficient of determination (R2) value of the Neural Network at 0.001 is higher than the R2 value of SVR, which reaches -0.036. Research find- ings indicate that the Artificial Neural Network model slightly outperforms the Support Vector Regression model, with lower prediction error rates and better ability to explain data variability.

Artificial Neural Network Model for Predicting Mechanical Strengths of Economical Ultra-High-Performance Concrete Containing Coarse Aggregates: Development and Parametric Analysis.

Ultra-high-performance concrete with coarse aggregates (UHPC-CA) has the advantages of high strength, strong shrinkage resistance and a lower production cost, presenting a broad application prospect in civil engineering construction. In view of the difficulty in establishing a mathematical model to accurately predict the mechanical properties of UHPC-CA, the back-propagation artificial neural network (BP-ANN) method is used to fully consider the various influential factors of the compressive strength (CS) and flexural strength (FS) of UHPC-CA in this paper. By taking the content of cement (C), silica fume (SF), slag, fly ash (FA), coarse aggregate (CA), steel fiber, the water-binder ratio (w/b), the sand rate (SR), the cement type (CT), and the curing method (CM) as input variables, and the CS and FS of UHPC-CA as output objectives, the BP-ANN model with three layers has been well-trained, validated and tested with 220 experimental data in the studies published in the literature. Four evaluating indicators including the determination coefficient (R2), the root mean square error (RMSE), the mean absolute percentage error (MAPE), and the integral absolute error (IAE) were used to evaluate the prediction accuracy of the BP-ANN model. A parametric study for the various influential factors on the CS and FS of UHPC-CA was conducted using the BP-ANN model and the corresponding influential mechanisms were analyzed. Finally, the inclusion levels for the CA, steel fiber, and the dimensionless parameters of the W/B and sand rate were recommended to obtain the optimal strength of UHPC-CA.

Web Application Based on MachineLearning for Diabetes Detection usingMicrostrip Resonator and Streamlit

Globally, the number of people living with diabetes is estimated to continue to increase. Blood needs to be drawn from the body to measure glucose. Invasive methods carry the risk of infection. Estimating the device using non-invasive methods can reduce the risks of invasive procedures. One way is to use machine learning. In this research, a machine learning-based web application was designed using the Artificial Neural Network (ANN) and K-nearest neighbor (KNN) methods to classify a person's diabetes type by adding additional features from the microstrip resonator. The data used is 1000 with 11 features and one output with two labels. The 11 features are gender, age, blood sugar levels, smoking habits, family history of diabetes, height, weight, Body Mass Index (BMI), frequency, return loss, and bandwidth. In the output, two labels are used, namely diabetes and non-diabetes. ANN and KNN both provide high accuracy above 90%. ANN delivers an accuracy of 99.3%, and KNN provides an accuracy of 99.67%. This model will be embedded in web applications created using Streamlit. It is hoped that this web application can simplify and speed up the diagnosis of diabetes so that the disease can be treated quickly and precisely from an early age. This is expected to minimize the dangerous effects of diabetes due to delays in treatment.

Integrated photonuclear cross sections in the giant dipole resonance of odd-mass actinide nuclei

This study explores the integrated total photonuclear cross section (σ0) within the context of the giant dipole resonance (GDR) in odd-mass actinide nuclei. Using artificial neural networks (ANNs) and adaptive neuro-fuzzy ınference system (ANFIS) machine learning algorithms, we analyze the GDR behaviors associated with the σ0 values in these nuclei. The modeling results obtained from ANFIS and ANN are compared among themselves and with the Translational Galilean Invariant Quasiparticle Phonon Nuclear Model (TGI-QPNM) and experimental data. Machine learning analysis and TGI-QPNM results provide valuable insights into the GDR characteristics of odd-mass actinides, shedding light on their photonuclear properties. The ANFIS model has achieved an R2 value of 0.98 and an RMSE of 0.19, while the ANN model (LM) has yielded an R2 value of 0.95 and an RMSE of 0.24. These findings deepen our understanding of nuclear physics, underscoring the role of artificial intelligence techniques in deciphering complex phenomena within nuclear structures. In conclusion, our study suggests that the ANFIS model, in agreement with TGI-QPNM results, generally outperforms the ANN (LM) method and could be a more effective tool for estimating the energy-weighted sum rule for GDR.

Artificial neural network aided vapor–liquid equilibrium model for multi-component high-pressure transcritical flows with phase change

In this work, an artificial neural network (ANN) aided vapor–liquid equilibrium (VLE) model is developed and coupled with a fully compressible computational fluid dynamics (CFD) solver to simulate the transcritical processes occurring in high-pressure liquid-fueled propulsion systems. The ANN is trained in Python using TensorFlow, optimized for inference using Open Neural Network Exchange Runtime, and coupled with a C++ based CFD solver. This plug-and-play model/methodology can be used to convert any multi-component CFD solver to simulate transcritical processes using only open-source packages, without the need of in-house VLE model development. The solver is then used to study high-pressure transcritical shock-droplet interaction in both two- and four-component systems and a turbulent temporal mixing layer (TML), where both qualitative and quantitative agreement (maximum relative error less than 5%) is shown with respect to results based on both direct evaluation and the state-of-the-art in situ adaptive tabulation (ISAT) method. The ANN method showed a 6 times speed-up over the direct evaluation and a 2.2-time speed-up over the ISAT method for the two-component shock-droplet interaction case. The ANN method is faster than the ISAT method by 12 times for the four-component shock-droplet interaction. A 7 times speed-up is observed for the TML case for the ANN method compared to the ISAT method while achieving a data compression factor of 2881. The ANN method also shows intrinsic load balancing, unlike traditional VLE solvers. A strong parallel scalability of this ANN method with the number of processors was observed for all the three test cases. Code repository for 0D VLE solvers, and C++ ANN interface—https://github.com/UMN-CRFEL/ANN_VLE.git.

Modeling of hydrogen separation through Pd membrane with vacuum pressure using Taguchi and machine learning methods

The performance of hydrogen purification using palladium (Pd) membrane is analyzed by Taguchi and machine learning (ML) methods. Three system factors are considered: the feed gas mixture composition (i.e., H2, CO2, and H2O), retentate-side total pressure, and vacuum pressure. The Taguchi method designs the experimental cases based on the constructed orthogonal array. In addition, the hydrogen flux is investigated by the analyses of variance (ANOVA) and artificial neural networks (ANN) methods. The results show that the effects of the considered factors on the hydrogen permeation performance can be ranked as feed gas mixture composition > retentate-side total pressure > vacuum pressure. Both the retentate-side pressure and vacuum pressure present a positive effect on hydrogen flux. The average relative error of hydrogen flux between the experimental results and predictions by ANN is 2.1%, which proves that the ANN method is an effective technique for predicting the hydrogen flux through the Pd membrane. The ML classification test is then performed for hydrogen purity by three machine learning methods: decision tree (DT), support vector machine (SVM), and ensemble method. Among the various classification models, the Quadratic SVM model exhibits a relatively high average training accuracy but shows overfitting. However, the bagged model of the ensemble method can achieve an impressive training accuracy of 91.4% and a prediction accuracy of 85.7%.

Design of a perturb and observe and neural network algorithms-based maximum power point tracking for a grid-connected photovoltaic system

Integrating photovoltaic systems (PV) into the grid has garnered significant attention due to increasing interest in renewable energy sources. Maximizing the PV systems output power is crucial for improving energy efficiency and reducing operating costs. This paper presents a comparative analysis of two different techniques of maximum power point tracking (MPPT): perturb and observe (P&O) and artificial neural network (ANN) MPPT, focusing on their application in grid-connected PV systems. The paper evaluates their performance under various operating conditions, including changes in irradiance and temperature, that are discussed in three cases. The comparative analysis includes metrics such as voltage regulation and powerloss. MATLAB Simulink is utilized to implement P&O and ANN MPPT methods, which include a PV cell connected to an MPPT-controlled boost converter. The simulation demonstrates the power loss of the PV model as well as the voltage regulation in the three cases for the two methods. The results obtained in simulation and implementations show that the ANN method outperforms the P&O in the three cases discussed in terms of powerloss, voltage regulation, and efficiency. The results also show that the change in output power from PV is noticeable when compared to changes in radiation, while the change is slight when temperatures change.