Application Of Artificial Neural Networks Research Articles

Application of artificial neural network, evolutionary polynomial regression, and life cycle assessment techniques to predict the performance of a new designed solar air ventilator with phase change material

Application of artificial neural network, evolutionary polynomial regression, and life cycle assessment techniques to predict the performance of a new designed solar air ventilator with phase change material

An integrated review on the role of different biocatalysts, process parameters, bioreactor technologies and data-driven predictive models for upgrading biogas.

An integrated review on the role of different biocatalysts, process parameters, bioreactor technologies and data-driven predictive models for upgrading biogas.

Artificial Neural Networks in Spray Coatings for Protection of Hydroturbines in Silt Conditions

Background: Artificial intelligence techniques are able to predict the erosion in various coatings. In recent times, the erosion of the coating was successfully predicted using artificial neural networks (ANN). Aim: This patent aims to present the critical findings in the area of spray coatings used for the protection of hydroturbines. This critical review explores the surface erosion of numerous hydroturbine materials and coatings. Moreover, it covers the evolution of coating techniques used in different industries such as hydropower plants and materials handling industry. Objective: The objective of this patent is to present the critical finding in the field of erosion wear in coatings utilizing an artificial neural networks (ANN) technique. Methods: The effect of influencing factors on silt erosion was assessed. However, the evolution of various coating processes was explored comprehensively. Furthermore, the properties of various coatings, namely WC, Cr2C3, and Ni/Co coatings, were reported. A number of comparative analyses were carried out that can help in the selection of coatings on the basis of their mechanical as well as chemical properties. The applications of ANN in the spray industry were explained extensively. Results: The results show that the HVOF coating is the most widely used deposition method. HVOF is an appropriate technique for both ceramics and pure metals. Ceramics, as well as cermet, are deposited to enhance the wear and tear resistance of hydroturbine materials. As a result, the HVOF, laser cladding, D-Gun, and plasma spray technique are optimal in these circumstances. Coatings based on WC have been employed by several researchers to increase stainless steels' resistance to erosion and wear. Conclusion: With the application of appropriate models and ML methods, it is expected that coatings with specific desired properties can be manufactured with fewer experiments. There is no surprise that WC-10Co-4Cr finds widespread usage in a variety of industrial contexts. Ni- and Cobased coatings show promise for use in tribological settings. Stellite is suggested by several researchers for high-erosion duty conditions in hydroturbines. Co-based coatings are more erosion resistant as compared to No-based coatings. Coatings formulated with chromium (Cr) are resistant to erosion and wear in harsh mining environments.

Intelligent Fault Diagnosis in 330 kV Power Networks Using SVM and ANN Techniques: Case of the Onitsha-New Haven Route

This paper investigated the application of Artificial Neural Networks (ANNs) and Support Vector Machines (SVMs) for transient fault detection and classification in power transmission systems, focusing on Nigeria’s Onitsha-New Haven 330 kV network. A Multilayer Perceptron (MLP) ANN, optimised with 10 hidden layers and the Levenberg-Marquardt algorithm, achieved near-perfect regression (correlation coefficient R > 0.999 ) and classification accuracy (≈99%) using 9,180 operational samples, validated against IEEE-compliant transient stability indices (Si) and dynamic voltage margins (DVm). Comparatively, SVMs demonstrated ~95–98% accuracy with sub-5 inference times, using pre-engineered features (such as harmonic distortions and wavelet coefficients) for real-time efficiency in resource-constrained environments. The study proposed a hybrid ANN-SVM framework, combining SVM’s rapid fault detection with ANN’s precision for post-event diagnostics, addressing the speed-accuracy trade-off in dynamic grids. Engineering implications highlight enhanced grid resilience, cost-effective deployment strategies, and support for renewable integration. At the same time, contributions include empirical validation of AI models in sparse-data contexts and methodological benchmarks aligning machine learning with power quality standards. Recommendations advocated phased AI adoption, workforce training, and regulatory standardisation, with future research directions spanning hardware-in-the-loop validation, cybersecurity, and adaptive learning for renewable-rich grids. This work bridges theoretical AI advancements with practical power system needs, offering scalable solutions for global energy transitions.

Rainfall-River Discharge Modelling Using Artificial Neural Network – A Case Study of Oramiriukwa River in Owerri, Imo State Nigeria

This study investigates the application of Artificial Neural Networks (ANNs) for rainfall-river modelling in the Oramiriukwa River, located in Owerri, Imo State, Nigeria. The study utilizes daily streamflow data from the Ulakwo station (1978–1988) alongside corresponding rainfall and temperature data for Imo State, obtained from the Nigerian Meteorological Agency (NiMet). A series of Feedforward Multi-Layer Perceptron (MLP) models were developed and tested using MATLAB, with performance evaluated using Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Coefficient of Determination (R²). Among the models, Model 4 ([15]) delivered the best results, achieving an R² of 0.9158, MSE of 0.1294, and RMSE of 0.3597, demonstrating its effectiveness for streamflow prediction in the Oramiriukwa River. Model 2 ([30, 15, 5]) also showed good performance (R² = 0.9029), but its increased complexity suggested a potential risk of overfitting. Model 1 ([20, 10]) yielded lower predictive accuracy, highlighting the need for more complex architectures or additional input features to improve ANN performance for hydrological applications. These results confirm the effectiveness of ANNs in modelling nonlinear hydrological processes and suggest their potential for improving streamflow prediction in similar river basins. This study contributes to the growing use of data-driven methods in hydrological modelling in Nigeria and offers a foundation for future work aimed at enhancing the accuracy and robustness of ANN models.

Predicting growth parameters of biofertilizer inoculated pepper, using root capacitance assessments and artificial neural networks in two soils.

Monitoring the root system plays an important role in understanding plant physiological processes; however, its assessment using non-destructive methods remains challenging. Here, we evaluate the utility of root capacitance (CR) as a practical indicator of root function and its relationship to plant growth parameters in Capsicum annuum L. To improve the accuracy of root function assessment, we applied artificial neural networks (ANN) as a novel data evaluation approach, comparing its predictive performance against multiple linear regression (MLR). Across two soil types (sandy and sandy loam), we applied multiple treatments ranging from microbial inoculants to wool pellet and inorganic nitrogen sources primarily to test whether CR could detect differences in root activity and biomass production under different conditions. We measured root dry biomass, shoot dry biomass, and leaf N content, treating these variables as independent predictors in a statistical framework. Multiple linear regression (MLR) initially showed strong relationship between CR and both root and shoot biomass in sandy soil, and between CR and total plant N content in sandy loam. However, an ANN model consistently outperformed MLR in predicting CR from plant physiological parameters, as evidenced by lower mean absolute error (MAE) in all treatments. These findings confirm that CR correlates strongly with plant growth parameters and can reliably distinguish the effects of different soil amendments even those with markedly different nutrient-release profiles.

Data Analytics by Python AI, to Understand Deviation of Minimum Temperature During Winter (Using Artificial Neural Network)

This paper is based on data analysis by Python programming language on Google collaborator platform, supported by the AI code, model used as neural network, to understand the deviation of minimum temperature for the period November to February in comparison with recorded normal. The source of data is online data collection platform ‘INDIA METEOROLOGICAL DEPARTMENT, PUNE’ and the online site ‘OGIMET’. The surface data for Alipore (42807) and the recorded normal temperature, as available there has been collected in csv file format, then uploading the csv file in Google collaborator platform, executed analysis by Python neural network model technique, to understand the predicted output, i.e. pattern of deviation of minimum temperature, based on analysis of big data, with minimum temperature data from 1969 to 2024. In this case the input column used is the column created with data of difference between minimum temperature and recorded normal for that month. The mod value of temperature difference for the four months, November to February filtered from this big data set has been used as filtered and scaled data, subjected to analyse to understand the future trend of this deviation. Model fit has been used with train-test split in 80%-20% ratio. The artificial neural network model actually resembles the structure of human brain, which can perform intelligent tasks similar to human brain by the process of learning by the machine. Like human brain a similar network is there with mapping between input and output, information propagated through different layers, using back propagation technique to learn the model to get best output with minimum loss. This technique resembles to the networking of human brain and neuron, where information is propagated through synaptic joints to react accordingly. In this paper prediction as well as analysis has been made with the application of artificial neural network with used model as LSTM supported by activation function and optimiser to learn the model for best fit output. The loss value of the model has also been verified to understand the success rate of the model

Integrating theoretical studies and computational modeling for predicting nanofiber diameter using artificial intelligence

The purpose of this study was to predict the diameter of PVA-Averrhoa bilimbi nanofibers using artificial intelligence approaches, such as response surface methodology (RSM) and artificial neural networks (ANN). The RSM and ANN models incorporated three independent variables: voltage, distance, and A. bilimbi concentration, with the dependent variable being the diameter of the PVA-A. bilimbi nanofibers. The RSM approach resulted in an R-squared value of 65.06%, while the ANN approach achieved an R-squared value of 99.98%. The novelty of this research lies in the innovative application of RSM and ANN for predicting the diameter of nanofibers composed of polyvinyl alcohol (PVA) and A. bilimbi. It is expected that this model will assist technicians, researchers, and practitioners in optimizing the variables that are essential to the development of these nanofibers. The integration of artificial intelligence with electrospinning technology holds the potential to develop automated and efficient nanofiber manufacturing processes.

Integration of Grab Methods with Artificial Neural Networks for Enhanced Decision-Making Systems

This study explores the architecture, development, and practical applications of artificial neural networks (ANNs), with advances in Gray Relational Analysis (GRA) techniques. ANNs are computational models designed to model biological neural systems, consisting of layers of interconnected neurons—i.e., input, hidden, and output layers—connected by synaptic weights. Operating through a connectionist approach, these networks effectively mimic the four basic functions of biological neurons: receiving input, integrating information, processing data, and generating output. Traditional ANNs, although powerful, face limitations in embedded systems due to their reliance on high-precision digital information transfer and the resulting resource demands. This has prompted the development of more efficient alternatives, such as spiking neural networks (SNNs), which use event-driven spiking signals to reduce power consumption and memory usage. The field has advanced further by incorporating theoretical models for quantum neural computing and the use of genetic algorithms employing various crossovers and mutation techniques. The versatility of ANNs is evident in a variety of applications. In healthcare, they aid in pattern recognition associated with conditions such as breast cancer and diabetes. For water resource management, ANN models predict relationships between rainfall and water levels. In financial sectors, they analyze complex economic conditions and assess credit risk for small business loans. Industrial applications include modelling complex systems in manufacturing plants, although widespread adoption in this field is limited. Complementing neural network advances, GRA methods have emerged to address multi-criteria decision-making challenges. Notable advances include the GRAS techniques have been developed to correct matrices with negative values, while fuzzy GRA methods now include interval-valued triangular fuzzy numbers and probabilistically uncertain linguistic word sets. Recent advances include innovations such as score values and normalized Hamming distances within single-valued neutrosophic fuzzy stein summary, these computational approaches offer powerful ways to solve complex, multidimensional problems, with recent studies highlighting improved performance and promising prospects for future application.

Integration of Grab Methods with Artificial Neural Networks for Enhanced Decision-Making Systems

This study explores the architecture, development, and practical applications of artificial neural networks (ANNs), with advances in Gray Relational Analysis (GRA) techniques. ANNs are computational models designed to model biological neural systems, consisting of layers of interconnected neurons—i.e., input, hidden, and output layers—connected by synaptic weights. Operating through a connectionist approach, these networks effectively mimic the four basic functions of biological neurons: receiving input, integrating information, processing data, and generating output. Traditional ANNs, although powerful, face limitations in embedded systems due to their reliance on high-precision digital information transfer and the resulting resource demands. This has prompted the development of more efficient alternatives, such as spiking neural networks (SNNs), which use event-driven spiking signals to reduce power consumption and memory usage. The field has advanced further by incorporating theoretical models for quantum neural computing and the use of genetic algorithms employing various crossovers and mutation techniques. The versatility of ANNs is evident in a variety of applications. In healthcare, they aid in pattern recognition associated with conditions such as breast cancer and diabetes. For water resource management, ANN models predict relationships between rainfall and water levels. In financial sectors, they analyze complex economic conditions and assess credit risk for small business loans. Industrial applications include modelling complex systems in manufacturing plants, although widespread adoption in this field is limited. Complementing neural network advances, GRA methods have emerged to address multi-criteria decision-making challenges. Notable advances include the GRAS techniques have been developed to correct matrices with negative values, while fuzzy GRA methods now include interval-valued triangular fuzzy numbers and probabilistically uncertain linguistic word sets. Recent advances include innovations such as score values and normalized Hamming distances within single-valued neutrosophic fuzzy stein summary, these computational approaches offer powerful ways to solve complex, multidimensional problems, with recent studies highlighting improved performance and promising prospects for future application.

Мониторинг деформаций перегонных туннелей метрополитена с применением нейросетевого подхода

The article discusses application of artificial neural networks and machine learning algorithms for monitoring of deformations in subway tunnels located in complex mining and geological conditions. Particular attention is given to industrial and environmental safety, as well as modern methods for measuring crustal deformations using GPS/GLONASS technologies, geodetic, and mine surveying. The main stages of artificial neural networks operation are described, i.e. the training based on the tunnel parameters and conditions, testing, validation, and operation to predict the potential deformations. The key neural network architectures are considered such as the deep, convolutional, and recurrent networks along with their data processing capabilities. Examples are provided of artificial neural networks used for data interpolation, hazardous zone recognition, and tunnel ring monitoring. The importance of high-quality initial data, including geometric parameters, physical material properties, climatic conditions, and historical data, is emphasized. Implementation of artificial neural networks can help to promptly identify risks, predict the deformation dynamics, and classify the deformation types, enabling timely measures to prevent emergencies.

Residual Temporal Convolutional Network With Dual Attention Mechanism for Multilead-Time Interpretable Runoff Forecasting.

As a pivotal subfield within the domain of time series forecasting, runoff forecasting plays a crucial role in water resource management and scheduling. Recent advancements in the application of artificial neural networks (ANNs) and attention mechanisms have markedly enhanced the accuracy of runoff forecasting models. This article introduces an innovative hybrid model, ResTCN-DAM, which synergizes the strengths of deep residual network (ResNet), temporal convolutional networks (TCNs), and dual attention mechanisms (DAMs). The proposed ResTCN-DAM is designed to leverage the unique attributes of these three modules: TCN has outstanding capability to process time series data in parallel. By combining with modified ResNet, multiple TCN layers can be densely stacked to capture more hidden information in the temporal dimension. DAM module adeptly captures the interdependencies within both temporal and feature dimensions, adeptly accentuating relevant time steps/features while diminishing less significant ones with minimal computational cost. Furthermore, the snapshot ensemble method is able to obtain the effect of training multiple models through one single training process, which ensures the accuracy and robustness of the forecasts. The deep integration and collaborative cooperation of these modules comprehensively enhance the model's forecasting capability from various perspectives. Ablation studies conducted validate the efficacy of each module, and through multiple sets of comparative experiments, it is shown that the proposed ResTCN-DAM has exceptional and consistent performance across varying lead times. We also employ visualization techniques to display heatmaps of the model's weights, thereby enhancing the interpretability of the model. When compared with the prevailing neural network-based runoff forecasting models, ResTCN-DAM exhibits state-of-the-art accuracy, temporal robustness, and interpretability, positioning it at the forefront of contemporary research.

Classification of Samples via Neural-Network Augmented Two-Dimensional Infrared Spectroscopy.

The application of artificial neural network (ANN) techniques to spectroscopy has proven to be a powerful tool for the rapid and accurate classification of experimental samples. However, despite the unique abilities of two-dimensional infrared spectroscopy (2D IR), the use of ANNs to classify samples on the basis of their 2D-IR spectra has been unexplored. We present two investigations into utilizing ANNs to perform end-to-end classification of samples from their 2D-IR spectra. In the first, we construct a model that can perform a binary classification of experimental samples on the basis of their solvent. In the second, we demonstrate that classification is possible even for a single spectral slice of pump-delay and waiting-time combination even when samples display almost identical spectra. These results clearly demonstrate the potential of ANN-augmented 2D IR, with particular emphasis on its use as a technology for high-throughput screening applications.

Application of Levenberg-Marquardt artificial neural network to study nanoparticle aggregation phenomena in stagnation point flow towards an off-centered rotating disk

Application of Levenberg-Marquardt artificial neural network to study nanoparticle aggregation phenomena in stagnation point flow towards an off-centered rotating disk

Application of principal component analysis and Artificial neural networks for the prediction of QoS in FSO links over South Africa

Application of principal component analysis and Artificial neural networks for the prediction of QoS in FSO links over South Africa

PERFORMANCE COMPARISON OF DIFFERENT ACTIVATION FUNCTIONS IN NEURAL NETWORKS FOR BIOMASS ENERGY CONTENT PREDICTION

The application of artificial neural networks to solve complex linear and nonlinear problems such as predicting biomass higher heating value (HHV) requires meticulous choice of the number of layers and neurons per layer, the training algorithms, and the activation functions among other hyper-parameters. Although some studies have examined these hyper-parameters, the effects of different activation functions on biomass HHV prediction have not attracted credible research attention. This study, therefore, employs three distinct activation functions (logsig, tansig, and purelin) in artificial neural networks for biomass HHV prediction based on proximate analysis. A 3-10-1 network architecture was used and the variation of the hidden layer and output layer activation functions yielded nine models (M1-M9) whose performances were assessed and compared using statistical indices. The results showed that the best performance was observed by model M2 which utlized the logsig function in the hidden layer and the tansig function in the output layer. This model had the highest determination coefficient of 0.8814 and the lowest mean square error and mean absolute error of 0.0017 and 0.0281 respectively. Understanding how these hyper-parameters influence biomass HHV prediction would guide the energy community to identify an optimal pathway to bioenergy production.

Application of artificial neural network for prediction of fracture energy of concrete

Application of artificial neural network for prediction of fracture energy of concrete

Prediction of AC Breakdown Voltage of Different Electrode Configuration Using Artificial Neural Networks

The study focuses on the application of Artificial Neural Networks (ANNs) to model and predict the AC breakdown voltage in gas-insulated systems for different electrode configurations. Gas-insulated systems play a crucial role in the transmission and distribution of electrical energy, and understanding the breakdown characteristics is essential for designing reliable and efficient systems. The research investigates how different electrode configurations impact the AC breakdown voltage in gas-insulated environments. Traditional methods for studying AC breakdown involve complex theoretical models and extensive experimentation. The use of ANN offers an innovative approach to predict breakdown voltages by learning complex relationships from available data.

Application of Artificial Neural Network in Web Expert System for Emotional Preschool Children

This study developed a web-based expert system to assist parents in identifying emotional parenting styles for preschool children using an Artificial Neural Network (ANN). The ANN model consists of an input layer, one hidden layer, and an output layer, using the sigmoid activation function. The parenting styles assessed include authoritarian, democratic, and permissive patterns. The observed symptoms refer to parental behaviors, such as the level of control, frequency of punishment, and involvement in conversations with the child. Data were collected through interviews with psychology experts and literature reviews, resulting in a dataset of 384 samples. The system was developed using the Waterfall software development model. Testing with ten sample cases showed that the system's predictions matched expert evaluations in nine cases (90% accuracy), while one case indicated a discrepancy. These results demonstrate that ANN can effectively classify parenting patterns based on observable symptoms. This expert system offers a practical tool for parents and educators to better understand and apply appropriate parenting strategies, contributing to the emotional development of preschool children.

Application of Artificial Neural Network for Prediction of Concrete–High-Performance Concrete Interfacial Bond Strength After Exposure to Elevated Temperature

High-performance concrete (HPC) is a new advanced building material for highway bridges, building construction, and repair/strengthen concrete structures with fire risk owing to its high fire resistance. The concrete composites should have interfacial bond strength (IBS) that is sufficient to transfer load between concrete components. When those composite structures are exposed to fire, horizontal cracks have been observed, and in some cases, the concrete layers have separated depending on the fire intensity. Therefore, the assessment of the IBS between the two concrete layers after exposure to fire is important for examining the entire fire behavior. Thus, the purpose of this work is to create an artificial neural network (ANN) model between statistically important factors and the IBS after exposure to elevated temperatures for using in the structural fire design of composite concrete layers. A total of 467 data points, including 252 data points from the slant shear test, 87 data points from the push-off test, and 128 data points from the tensile test, have been collected from literature reviews. Firstly, the independent parameters such as interfacial surface roughness, temperature exposure, part of the specimen exposed to temperature, type of concrete overlay, and fiber content introduced in the concrete overlay were carefully analyzed to identify the statistically important parameters and their impact on the IBS. Secondly, a designed ANN model has been developed to predict the IBS based on the type of test technique, interfacial surface roughness, temperature exposure, type of concrete overlay, and fiber content. Moreover, a mathematical model has been proposed to predict the IBS between concrete substrate and HPC after exposure to elevated temperature. Finally, the predicted IBS from the design ANN model and the mathematical IBS were compared with the available empirical models from literature. The outcome results demonstrated that the design ANN model was able to predict the IBS between two concrete layers after exposure to elevated temperatures with a coefficient of determination R2 of 0.97, while the mathematical IBS gave a good accuracy for predicting the IBS in the case of the interface under combined stress with R2 equal to 0.90. This study effectively bridges the gaps in both theoretical and experimental findings by integrating ANN models with advanced computational techniques and robust statistical analyses. This multifaceted approach not only enriches our understanding of the topic, but also provides more precise insights and predictive capabilities.