Basis Of Artificial Neural Network Research Articles

AOSMA-MLP: A Novel Method for Hybrid Metaheuristics Artificial Neural Networks and a New Approach for Prediction of Geothermal Reservoir Temperature

To ascertain the optimal and most efficient reservoir temperature of a geothermal source, long-term field studies and analyses utilizing specialized devices are essential. Although these requirements increase project costs and induce delays, utilizing machine learning techniques based on hydrogeochemical data can minimize losses by accurately predicting reservoir temperatures. In recent years, applying hybrid methods to real-world challenges has become increasingly prevalent over traditional machine learning methodologies. This study introduces a novel machine learning approach, named AOSMA-MLP, integrating the adaptive opposition slime mould algorithm (AOSMA) and multilayer perceptron (MLP) techniques, specifically designed for predicting the reservoir temperature of geothermal resources. Additionally, this work compares the basic artificial neural network and widely recognized algorithms in the literature, such as the whale optimization algorithm, ant lion algorithm, and SMA, under equal conditions using various evaluation regression metrics. The results demonstrated that AOSMA-MLP outperforms basic MLP and other metaheuristic-based MLPs, with the AOSMA-trained MLP achieving the highest performance, indicated by an R2 value of 0.8514. The proposed AOSMA-MLP approach shows significant potential for yielding effective outcomes in various regression problems.

Analysis of Properties of Polymer Composites with Various Types of Fillers

The processes of structure formation of composite building materials (KSM) on different polymer binders are presented. It is shown that one of the most significant components of KSM are fillers, which help to improve their structural and operational characteristics. This work is devoted to the analysis of the results of an experimental study of the properties of epoxy composites with fillers having various elastic-plastic and strength properties. The research was carried out in three stages: at the first stage, studies were conducted aimed at assessing the influence of the nature of the filler on the curing processes of KSM; at the second, the influence of the type of filler and its quantitative content on the strength of composites was established, at the third, compositions were optimized using fillers with different indicators of grain composition and elastic-plastic properties. Powders of glass, dolomite, thermolite, and diatomite were considered as fillers at the first and second stages of the research, and powders of glass, ceramics, and chalk were considered at the second stage. The research at the third stage was carried out using mathematical methods of experiment planning with the construction of a planning matrix for a complete factor experiment and the determination of the values of the response functions relative to the encoded factors. The physico-mechanical properties, degree of curing, and chemical resistance of filled epoxy CCM have been established. On the basis of artificial neural networks, the maximum properties of the studied composites with fillers were determined. An assessment of structural properties based on rank correlation is also proposed. The results of the research can be used to predict the properties of KSM, as well as to clarify the extreme parameters of the properties. The dependences of changes in the properties of polymer composites on the surface characteristics, the dispersion of fillers and the degree of filling were established; preferred fillers for epoxy composites were determined; fillers were determined to assess the effect of elastic surface properties of composites, allowing to improve the strength and deformability of polymer composites; regression models were obtained based on a complete factorial experiment; an assessment of the «structural stability» of the studied composites using Pearson, Kendall, Spearman rank correlation; On the basis of artificial neural networks, the extreme properties of the studied composites with fillers were determined, neural networks.

Neuro – symbolic AI for materials modelling and processes design

This paper presents possibilities of combined use of Artificial Neural Networks (ANNs) with symbolic AI methods for improving the creation and the efficiency of models for materials and manufacturing processes. Besides a short presentation of one of the most basic and common ANNs architecture, which is the Multi-Layer Perceptron (MLP), and the presentation of main symbolic AI methods for knowledge representation and processing, the paper discusses their combined use and possibilities of hybrid neuro–symbolic AI applications for modelling materials and processes. This combined AI modelling approach is planned to be implemented and tested within the Horizon Europe project M2DESCO, which concerns modelling and eco-design of High Entropy Alloys’ (HEAs) coatings.

Machine learning based post‐processing of model‐derived near‐surface air temperature – A multimodel approach

AbstractIn this article, a machine‐learning‐based tool for calibrating numerical forecasts of near‐surface air temperature is proposed. The study area covers Poland representing a temperate type of climate with transitional features and highly variable weather. The direct output of numerical weather prediction (NWP) models is often biased and needs to be adjusted to observed values. Forecasters have to reconcile forecasts from several NWP models during their operational work. As the proposed method is based on deterministic forecasts from three short‐range limited‐area models (ALARO, AROME and COSMO), it can support them in their decision‐making process. Predictors include forecasts of weather elements produced by the NWP models at synoptic weather stations across Poland and station‐embedded data on ambient orography. The Random Forests algorithm (RF) has been used to produce bias‐corrected forecasts on a test set spanning one year. Its performance was evaluated against the NWP models, a linear combination of all predictors (multiple linear regression, MLR) as well as a basic Artificial Neural Network (ANN). Detailed evaluation was done to identify potential strengths and weaknesses of the model at the temporal and spatial scale. The value of RMSE of a forecast obtained by the RF model was 11% and 27% lower compared to the MLR model and the best‐performing NWP model respectively. The ANN model turned out to be even superior, outperforming RF by around 2.5%. The greatest improvement occurred for warm bias during the nighttime from July to September. The largest difference in forecast accuracy between RF and ANN appeared for temperature drops inApril nights. Poor performance of RF for extreme temperature ranges may be suppressed by training the model on forecast error instead of observed values of the variable.

Predictive modelling framework on the basis of artificial neural network: A case of nano-powder mixed electric discharge machining

In this modern era where Industry 4.0, plays a crucial role in enhancing productivity, quality, and resource utilization by digitalizing and providing smart operation to industrial systems. Therefore, there is a need to establish a framework that enhances productivity and quality of work to achieve the net-zero from industry. In this study, a comprehensive and generic analytical framework has been established to mitigate or lessen the research and technological gap in the manufacturing sector. In addition to that, the key stages involved in artificial intelligence (AI) based modelling and optimization analysis for manufacturing systems have also been incorporated. To assess the proposed AI framework, electric discharge machining (EDM) as a case study has been selected. The focus enlightens the emergence of optimizing the material removal rate (MRR) and surface roughness (SR) for Inconel 617 (IN617) material. A full factorial design of the experiment was carried out for experimentation. After that, an artificial neural network (ANN) as a modelling framework is selected, and fine-tuning of hyperparameters during training has been carried out. To validate the predictive performance of the trained models, an external validation (Valext) test has been conducted. Through sensitivity analysis (SA) on the developed AI framework, the most influential factors affecting MRR and SR in EDM have been identified. Specifically, powder concentration (Cp) contributes the most to the percentage significance, accounting for 79.00 % towards MRR, followed by treatment (16.35 %) and 4.67 % surfactant concentration (Sc). However, the highest % significance in SR is given by Sc (36.86 %), followed by Cp (33.23 %), and then treatment (29.90 %), respectively. Furthermore, a parametric optimization has been performed using the framework and found that MRR and SR are 93.75 % and 58.90 % better than experimental data. This successful performance optimization proposed by the framework has the potential for application to other manufacturing systems.

Smart reference evapotranspiration using Internet of Things and hybrid ensemble machine learning approach

Reference Evapotranspiration (ETo) is the cornerstone of efficient water utilization for sustainability in agriculture. The standard Penman–Montieth (PM) approach of Reference Evapotranspiration (ETo), is complex due to the involvement of an extensive set of climatic conditions. The existing solutions of simplification of ETo predictions are not in accordance with the Penman–Montieth approach. A hybrid ensemble machine learning approach for simplification of ETo prediction is proposed using the Internet of Things(IoT) based crop field sensed climatic data. The proposed hybrid ensemble model is implemented with an Artificial Neural Network (ANN) and regression models. The proposed solution is unique for its utilization of flexible climatic conditions and in accordance with the standard Penman–Montieth (PM) approach. The proposed solution is able to predict daily ETo from only temperature and also can adjust ETo according to wind speed, humidity, and sunshine duration. The assessment of the proposed model exhibits a high coefficient of determination (R2) of 0.94 compared to 0.91 from the basic ANN model. The proposed hybrid ensemble model also exhibits a low RMSE of 0.86, MAE of 0.75 mm day−1, and MAPE of 15.05%, compared to 0.91, 0.75 mm day−1, and 20.40% from the basic ANN model. The ETo predictions by the proposed hybrid ensemble model also exhibit a higher Pearson correlation coefficient of 0.917 with the ETo by the Penman–Montieth (PM) approach, compared to 0.778 by the basic ANN model. The statistics reveal the accuracy and goodness of fit of the proposed hybrid ensemble machine learning model.

Moving Target Detection Algorithm for Dynamic Image Sequences on The Basis of Artificial Neural Network

<p>Due to the large changes in dynamic image sequence frames and the complex detection scene, it is difficult to accurately detect moving objects. Therefore, the study proposes a moving target detection algorithm based on artificial neural network. First, the algorithm performs standardized grayscale processing and gamma correction processing on the dynamic image to eliminate the noise interference of the dynamic image. After that, the model calculates the gradient of the dynamic image in order to complete the feature extraction of the dynamic image. Then, according to the result of hog feature extraction, the study adopts the inter-frame calculation method to update the background of the dynamic image. Finally, the principle and structure of the neural network are analyzed experimentally, and a channel attention mechanism is introduced to train dynamic image sequences to obtain MTD results. Experimental results show that the proposed algorithm achieves higher accuracy in MTD than conventional detection algorithms. The calculation efficiency of the algorithm in this paper has significant advantages, and the average detection time is 3.69515ms, which can meet the real-time requirements of MTD.</p> <p> </p>

Elements of the theory of designing manipulators from composite materials for ground robotic complexes for special purposes

Elements of the theory of designing manipulators from composite materials intended for use in ground-based robotic complexes for special purposes have been developed. Circuit solutions of manipulators and constructive execution of their main components and elements are proposed. Special hinges and remote drives for moving links of manipulators have been developed. To ensure the required static and dynamic characteristics, special mechatronic modules are implemented in the form of geared motors that are outside the manipulators. Control systems of geared motors are implemented on the basis of artificial neural networks using the results of mathematical modeling of statics and dynamics of manipulators by helical calculus methods. The developed manipulators significantly increase the functionality of ground robotic complexes when they work with dangerous objects.

Application of Machine Learning for Prediction and Process Optimization—Case Study of Blush Defect in Plastic Injection Molding

Injection molding is one of the most important processes for the mass production of plastic parts. In recent years, many researchers have focused on predicting the occurrence and intensity of defects in injected molded parts, as well as the optimization of process parameters to avoid such defects. One of the most frequent defects of manufactured parts is blush, which usually occurs around the gate location. In this study, to identify the effective parameters on blush formation, eight design parameters with effect probability on the influence of this defect have been investigated. Using a combination of design of experiments (DOE), finite element analysis (FEA), and ANOVA, the most significant parameters have been identified (runner diameter, holding pressure, flow rate, and melt temperature). Furthermore, to provide an efficient predictive model, machine learning methods such as basic artificial neural networks, their combination with genetic algorithms, and particle swarm optimization have been applied and their performance analyzed. It was found that the basic artificial neural network (ANN), with an average accuracy error of 1.3%, provides the closest predictions to the FEA results. Additionally, the process parameters were optimized using ANOVA and a genetic algorithm, which resulted in a significant reduction in the blush defect area.

Pumps-as-Turbines’ (PaTs) performance prediction improvement using evolutionary artificial neural networks

Energy production from clean sources is mandatory to reduce pollutant emissions. Among different options for hidden hydropower potential exploitation, Pump-as-Turbine (PaT) represents a viable solution in pico- and micro-hydropower applications for its flexibility and low-cost. Pumps are widely available in the global market in terms of both sizes and spare parts. To date, there are several PaTs’ performance prediction models in the literature, but very few of them use optimization algorithms and only for specific and limited prediction goals. The present work proposes evolutionary Artificial Neural Networks (ANNs) based on JADE, which is a typology of differential evolution algorithm, to forecast Best Efficiency Point (BEP) and performance curves of a PaT starting from the pump operational data. In this model, JADE is employed as optimizer of basic ANNs to upgrade parameter values of the learning rate, weights, and biases. The accuracy of the proposed model is evaluated through experimental data available from the literature and compared to a basic ANN and two versions of the differential evolution algorithm. Results are also validated with experiments on a PaT showing that the proposed method can achieve an average R2-value of 0.97, which is 5% higher than the one obtained with a basic ANN.

Intelligent Simulation of Water Temperature Stratification in the Reservoir.

In order to fully make use of limited water resources, humans have built many water conservancy projects. The projects produce many economic benefits, but they also change the natural environment. For example, the phenomenon of water temperature stratification often occurs in deep reservoirs. Thus, effective ways are needed to predict the water temperature stratification in a reservoir to control its discharge water temperature. Empirical formula methods have low computational accuracy if few factors are considered. Mathematical model methods rely on large amounts of accurate hydrological data and cost long calculation times. The purpose of the research was to simulate water temperature stratification in a reservoir by constructing an intelligent simulation model (ISM-RWTS) with five inputs and one output, determined on the basis of artificial neural networks (ANN). A 3D numerical model (3DNM) was also constructed to provide training samples for the ISM-RWTS and be used to test its simulation effect. The ISM-RWTS was applied to the Tankeng Reservoir, located in the Zhejiang province of China, and performed well, with an average error of 0.72 °C. Additionally, the Intelligent Computation Model of Reservoir Water Temperature Stratification (ICM-RWTS) was also discussed in this paper. The results indicated that the intelligent method was a powerful tool to estimate the water temperature stratification in a deep reservoir. Finally, it was concluded that the advantages of the intelligent method lay in its simplicity of use, its lower demand for hydrological data, its well generalized performance, and its flexibility for considering different input and output parameters.

Predicting the fracture mechanics responses of recycled asphalt mixes using machine learning-based algorithms: Application of CART algorithm and neural networks

This research was aimed at presenting a comprehensive comparison between the numerous asphalt fracture indices in terms of significance and applicability. Accordingly, recycled WMA mixes were assessed by conducting fracture testing at low-temperature conditions. Analytical analyses, probabilistic assessment, machine learning algorithms, and neural networks modeling were further investigated in responding to the research requirements. Three different categories of WMA with varied RAP contents (0, 30, 50, and 70 %) were used for mix preparation. Exposure to low temperatures and freeze–thaw cycles were the factors applied for specimen conditioning before conducting semi-circular bending tests. Analytical results showed that the inclusion of RAP particles in WMA mixes played a protective role when freeze–thaw cycles proceeded; particularly, the inclusion of 30 % of RAP particles showed comparable cracking performance with mixes without RAP. On the other hand, statistical analyses showed the promises behind of WeibullCRI index, compared with other fracture indices, in characterizing the cracking potential of recycled WMA mixes. Based on the classification and regression tree (CART) algorithm, temperature and RAP contents were highlighted as the most effective factors that governed the fracture properties of examined mixes. Results of machine learning-based analyses (CART algorithm) were promisingly consistent with what was observed from the probabilistic data analysis. Among the assessed factors and those of adopted levels, effects of freeze–thaw cycles (up to 3) and mix technology were noticed to be virtually negligible. Basic artificial neural networks and radial basis function neural networks were assessed based on comparing their efficiency in data training and prediction. The results showed that the radial basis function neural networks will promisingly predict the WeibullCRI values comprising specific model regulations.

Identification of Early Esophageal Cancer by Semantic Segmentation.

Early detection of esophageal cancer has always been difficult, thereby reducing the overall five-year survival rate of patients. In this study, semantic segmentation was used to predict and label esophageal cancer in its early stages. U-Net was used as the basic artificial neural network along with Resnet to extract feature maps that will classify and predict the location of esophageal cancer. A total of 75 white-light images (WLI) and 90 narrow-band images (NBI) were used. These images were classified into three categories: normal, dysplasia, and squamous cell carcinoma. After labeling, the data were divided into a training set, verification set, and test set. The training set was approved by the encoder–decoder model to train the prediction model. Research results show that the average time of 111 ms is used to predict each image in the test set, and the evaluation method is calculated in pixel units. Sensitivity is measured based on the severity of the cancer. In addition, NBI has higher accuracy of 84.724% when compared with the 82.377% accuracy rate of WLI, thereby making it a suitable method to detect esophageal cancer using the algorithm developed in this study.

Soft computing techniques for forecasting of COVID-19 in Pakistan

Novel Pandemic COVID-19 led globally to severe health barriers and financial issues in different parts of the world. The forecast on COVID-19 infections is significant. Demeanor vital data will help in executing policies to reduce the number of cases efficiently. Filtering techniques are appropriate for dynamic model structures as it provide reasonable estimates over the recursive Bayesian updates. Kalman Filters, used for controlling epidemics, are valuable in knowing contagious infections. Artificial Neural Networks (ANN) have generally been used for classification and forecasting problems. ANN models show an essential role in several successful applications of neural networks and are commonly used in economic and business studies. Long short-term memory (LSTM) model is one of the most popular technique used in time series analysis. This paper aims to forecast COVID-19 on the basis of ANN, KF, LSTM and SVM methods. We applied ANN, KF, LSTM and SVM for the COVID-19 data in Pakistan to find the number of deaths, confirm cases, and cases of recovery. The three methods were used for prediction, and the results showed the performance of LSTM to be better than that of ANN and KF method. ANN, KF, LSTM and SVM endorsed the COVID-19 data in closely all three scenarios. LSTM, ANN and KF followed the fluctuations of the original data and made close COVID-19 predictions. The results of the three methods helped significantly in the decision-making direction for short term strategies and in the control of the COVID-19 outbreak.

Forecasting of WTI crude oil using combined ANN-Whale optimization algorithm

ABSTRACT The current study predicts West Texas Intermediate (WTI) petroleum prices using an artificial neural network (ANN) with a whale optimization algorithm (WOA). In implementing the model, five parameters, including gold price, coal price, natural gas price, Dollar-Euro exchange rate, and Dollar-Yuan exchange rate, have been used as input to the combined model. The intelligent and basic ANN algorithm results compared to finding the ANN-WOA algorithm capacity in predicting the future price of WTI oil. ANN-WOA model improved the WTI price predicting accuracy up to 22% compared to the ANN. The ANN-WOA method with a value of R2 = 0.93 compared to the ANN method with a value of R2 = 0.75 was able to reduce the model error well. According to the significant impact that the input parameters of the combination model had on the WTI oil price prediction, therefore, in studies that predict price or other variables, highly correlated variables can significantly increase the accuracy of the forecast.

Information System to Analyze Human Environmental Wellbeing for Assessing and Reducing Technogenic Risk

The article provides a rationale for the need of developing the information system to analyze human environmental wellbeing (HEW) on the basis of artificial neural networks. The application of artificial neural networks is pre-conditioned by the need of prompt analysis and processing to take decisions on the implementation of environment protection measures for large data arrays accumulated as a result of environmental monitoring. The suggested information system for technogenic risk assessment implements a multi-limit way to control the state of a multi-parametric object and allows assessing human environmental wellbeing in real time. The authors provide structural charts for the information system to assess technogenic risk and the block to process and store information on the basis of an artificial neural network LVQ (Learning Vector Quantization). The authors provide an algorithm and the results for the artificial neural network LVQ training.

Automation of the monitoring the technical condition of ladle cars on the basis of artificial neural networks

Automation of the monitoring the technical condition of ladle cars on the basis of artificial neural networks

Neural Network Model for Control of Operating Modes of Crushing and Grinding Complex

This article investigates the application of neural network models to create automated control systems for industrial processes. We reviewed and analysed works on dispatch control and evaluation of equipment operating modes, and the use of artificial neural networks to solve problems of this type. It is shown that the main requirements for identification models are the accuracy of estimation and ease of algorithm implementation. It is shown that artificial neural networks meet the requirements for accuracy of classification problems, ease of execution and speed. We considered the structures of neural networks that can be used to recognize the modes of operation of technological equipment. Application of the model and structure of networks with radial basis functions and multilayer perceptrons for the tasks of identifying the mode of operation of equipment under given conditions is substantiated. The input conditions for the construction of neural network models of two types with a given three-layer structure are offered. The results of training neural models on the model of a multilayer perceptron and a network with radial basis functions are presented. The estimation and comparative analysis of models depending on model parameters is made. It is shown that networks with radial basis functions offer greater accuracy in solving identification problems. The structural scheme of the automated process control system with mode identification on the basis of artificial neural networks is offered.

An Artificial Neural Network-Based Model Predictive Control for Three-Phase Flying Capacitor Multilevel Inverter

Model predictive control (<i>MPC</i>) has been used widely in power electronics due to its simple concept, fast dynamic response, and good reference tracking. However, it suffers from parametric uncertainties, since it directly relies on the mathematical model of the system to predict the optimal switching states to be used at the next sampling time. As a result, uncertain parameters lead to an ill-designed <i>MPC</i>. Thus, this paper offers a model-free control strategy on the basis of artificial neural networks (<i>ANNs</i>), for mitigating the effects of parameter mismatching while having a little negative impact on the inverter&#x2019;s performance. This method includes two related stages. First, <i>MPC</i> is used as an expert to control the studied converter in order to provide a dataset, while, in the second stage, the obtained dataset is utilized to train the proposed <i>ANN</i>. The case study herein is based on a four-level three-cell flying capacitor inverter. In this study, MATLAB/Simulink is used to simulate the performance of the proposed method, taking into account various operating conditions. Afterward, the simulation results are reported in comparison with the conventional <i>MPC</i> scheme, demonstrating the superior performance of the proposed control strategy in terms of robustness against parameters mismatch and low total harmonic distortion (<i>THD</i>), especially when changes occur in the system parameters, compared to the conventional <i>MPC</i>. Furthermore, the experimental validation of the proposed method is provided based on the Hardware-in-the-Loop (HIL) simulation using the C2000TM-microcontroller-LaunchPadXL TMS320F28379D kit, demonstrating the applicability of the <i>ANN</i>-based control strategy to be implemented on a DSP controller.

Application of artificial intelligence in diagnosis of pancreatic malignancies by endoscopic ultrasound: a systemic review.

Background:Pancreatic cancer (PC) is a highly fatal malignancy with a global overall 5-year survival of under 10%. Screening of PC is not recommended outside of clinical trials. Endoscopic ultrasonography (EUS) is a very sensitive test to identify PC but lacks specificity and is operator-dependent, especially in the presence of chronic pancreatitis (CP). Artificial Intelligence (AI) is a growing field with a wide range of applications to augment the currently available modalities. This study was undertaken to study the effectiveness of AI with EUS in the diagnosis of PC.Methods:Studies from MEDLINE and EMBASE databases reporting the AI performance applied to EUS imaging for recognizing PC. Data were analyzed using descriptive statistics. The Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool was used to assess the quality of the included studies.Results:A total of 11 articles reported the role of EUS in the diagnosis of PC. The overall accuracy, sensitivity, and specificity of AI in recognizing PC were 80–97.5%, 83–100%, and 50–99%, respectively, with corresponding positive predictive value (PPV) and negative predictive value (NPV) of 75–99% and 57–100%, respectively. Types of AI studied were artificial neural networks (ANNs), convolutional neural networks (CNN), and support vector machine (SVM). Seven studies using other than basic ANN reported a sensitivity and specificity of 88–96% and 83–94% to differentiate PC from CP. Two studies using SVM reported a 94–96% sensitivity, 93%–99% specificity, and 94–98% accuracy to diagnose PC from CP. The reported sensitivity and specificity of detection of malignant from benign Intraductal Papillary Mucinous Neoplasms (IPMNs) was 96% and 92%, respectively.Conclusion:AI reported a high sensitivity with high specificity and accuracy to diagnose PC, differentiate PC from CP, and differentiate benign from malignant IPMN when used with EUS.