Artificial Neural Network Structure Research Articles

Hybrid ECBO–ANN Algorithm for Shear Strength of Partially Grouted Masonry Walls

In recent years, artificial neural network (ANN) has become one of the popular and effective machine learning models, having a unique ability to handle very complex problems and the potential to predict accurate results without a defined algorithmic solution. However, the ANN structure and parameters are usually chosen by experience.The behavior of Partially Grouted (PG) masonry shear walls is complex due to the inherent anisotropic properties of the masonry materials and the nonlinear interactions between mortar, blocks, grouted cells, non-grouted cells, and reinforcing steel.In this study, the aim is to develop an artificial neural network model by combining the ECBO meta-heuristic algorithm with the artificial neural network structure to optimize the feed forward propagation network parameters for analyzing the shear strength of PG walls.A total of 255 test data on PG collected from the available literature were used to generate training and test data sets. Various validation criteria such as mean square error, root mean square error and correlation coefficient (R) are used to validate the models.In this study, the optimal number of neurons used in the hidden layer and also the optimal number of CBs required in the ECBO algorithm were obtained. The mathematical formulation of the optimized neural network model with the combination of meta-heuristic algorithm is also presented.

Modeling of the Drag Force in Polydisperse Gas–Solid Flow via an Efficient Supervised Machine Learning Approach

Most granular flow in nature and industrial processing has the property of polydispersity, whereas we are always restricted to using the monodisperse drag force model in simulations since the drag force model with polydispersity is difficult to establish. Ignoring polydispersity often results in obvious deviations between simulation and experimental outcomes. Generally, it is very hard for us to describe the characteristics of polydispersity in drag force by using a function with analytic expression. Recently, the artificial neural network (ANN) model provides us the advantages of estimating these kinds of outcomes with better accuracy. In this work, the ANN is adopted to model the drag force in polydisperse granular flows. In order to construct a reasonable ANN algorithm for modeling the polydisperse drag force, the structures of ANN are elaborately designed. As training for the ANN drag model, a direct numerical simulation method is proposed, based on the lattice Boltzmann method (LBM), to generate the training data, and an adaptive data filtering algorithm, termed as the optimal contribution rate algorithm (OCRA), is introduced to effectively improve the training efficiency and avoid the over-fitting problems. The results support that the polydispersity of the system can be well scaled by the ANN drag model in a relatively wide range of particle concentrations, and the predicted results coincide well with the experimental ones. Moreover, the ANN drag model is not only effective for polydisperse systems, but compatible with monodisperse systems, which is impossible using traditional drag models.

Estimation of Alfalfa (Medicago sativa l.) yield under RCP4.5 and RCP8.5 climate change projections with ANN in Turkey

Alfalfa is one of the most widely cultivated forage crops in the world. Alfalfa farming is carried out on approximately 35 million ha of land worldwide with an annual production amounting to 255 million tons. The average alfalfa cultivated area is about 637 000 ha with production of 13 million tons and yield of 2 200 kg da-1 in Turkey. It is expected that climate change will have significantly different effects on its production and yield in future. Therefore, the aim of the study was to predict the effect of climate change on the yield of alfalfa via selected Artificial Neural Network (ANN) according to RCP4.5 and RCP8.5 climate change scenarios. In line with this, first of all the best ANN structure among 176 different ANN alternatives consisting of various input parameters, learning rates, decay and neuron numbers to predicts alfalfa yield was selected. The ANN training/test dataset used in the study were composed of the alfalfa cultivation statistics, the soil parameters and the climatological data. Alfalfa yield for years 2020-2060 and 2060-2100 in 79 provinces of Turkey are predicted by using best ANN model, according to climate change projections (HadGEM2-ES RCP4.5 and RCP8.5). The ANN was able to calculate alfalfa yield with 0.827 coefficient of determination and 0.813 Nash-Sutcliff coefficient. It is understood that the alfalfa plant can resist climate change and its yield tend to increase or decrease in regions, where there will be an increase or decrease in precipitation in the same order as result of climatic change. It is predicted that the highest yield increase will be in Artvin (6%) (a province of the Eastern Anatolia region) and the maximum yield decrease will be noted in Siirt (9%) (a province of the South eastern Anatolia region). This research may be considered as a creative prediction approach for the alfalfa yield estimation.

Control for Grid-Connected VSC With Improved Damping Based on Physics-Informed Neural Network

The rapid penetration of renewable energy sources into the power system makes the grid-connected voltage source converter (VSC) highly dynamic and uncertain. This necessitates designing new adaptive control for VSCs to ensure satisfactory system performance, reliability, and stability. This paper introduces a physics-informed artificial neural network (ANN) controller for the grid-connected VSC to improve the system performance and dampen the voltage oscillation due to the sudden change in power demand. The employed ANN structure is a feed-forward multilayer Neural Network trained offline by the Levenberg-Marquardt-based backpropagation algorithm. Results are presented for different dynamic scenarios to show the satisfactory operation of the proposed controller. The small-signal stability analysis is presented to validate the system's stability. Further, the performance of the proposed ANN controller is compared with the widely-used PI-controller and model predictive controller. The results prove that the proposed controller has a better dynamic performance in damping the voltage oscillation.

A magnetic-field-driven neuristor for spiking neural networks

Artificial intelligence has been widely deployed in many fields with remarkable success. Among various artificial neural network structures in artificial intelligence, the spiking neural network, as the next-generation artificial neural network, closely mimics the natural neural networks. It contains the all-or-nothing and diverse periodic spiking, which is an analogy to the behavior of natural neurons. Artificial devices that perform the function of neurons are called neuristors. Most existing neuristors are driven by electrical signals, which suffer the problem of impedance mismatch between input and output neuristors. By exploiting the S-shape negative differential resistances element that is sensitive to the external magnetic field, we constructed a magnetic-field-driven neuristor. Magnetic fields can stimulate all-or nothing spiking, and its shape and frequency can be modulated through capacitances in the circuit. As magnetic fields serve as the information carrier, the cascading of our neuristors can get rid of the electrical impedance mismatch, promising a scalable hardware platform for spiking neural networks.

A new study on the prediction of the effects of road gradient and coolant flow on electric vehicle battery power electronics components using machine learning approach

Transistor and diode junction temperature is an important factor in battery electric vehicle systems. High junction temperatures can lead to increased power losses and reduce the efficiency of the power conversion process. Additionally, high junction temperatures can accelerate the degradation rate of lithium-ion batteries, leading to capacity loss, increased internal resistance, and reduced overall battery life. In this study, the parameters affecting the transistor and diode junction temperatures of the battery group, which is the most important component of a battery electric vehicle, have been extensively analyzed and estimated by the machine learning approach. An artificial neural network (ANN) model has been established to estimate the parameters affecting the transistor and diode junction temperatures. The Bayesian regularization algorithm is preferred as the training algorithm in the ANN structure with multilayer perceptron architecture. There are 20 neurons in the hidden layer of the neural network established with a total of 242 data sets obtained from the literature. 80 % of the data set was employed for the training of the model and 20 % for the testing phase. The ANN model's calculated performance parameters were 8.81E-01 for mean squared error (MSE) and 0.91739 for coefficient of determination (R) value. The proposed ANN model was able to successfully predict the parameters affecting the transistor and diode junction temperatures with an average proportional deviation of less than 0.038 % from the actual values. This study can be considered a pioneering study both in terms of cost and time for production practice and in terms of its contribution to the literature.

A hybrid approach for forecasting bitcoin series

Bitcoin price prediction is a substantial challenge for cryptocurrency investors. This study offers an innovative scheme to predict Bitcoin returns and volatilities using a hybrid model that incorporates the autoregressive fractionally integrated moving average (ARFIMA), empirical wavelet (EW) transform, and local linear wavelet neural network (LLWNN) approaches to produce an ARFIMA-EWLLWNN model. Our methodologies integrate the advantages of the long-memory model, EW decomposition technique, artificial neural network structure, and backpropagation and particle swarm optimization learning algorithms. The experimental results of the optimized hybrid approach outperform some classic models by providing more accurate out-of-sample forecasts over longer horizons. The model proves to be the most appropriate Bitcoin forecasting technique. Moreover, the implemented method produces smaller prediction errors than other computing techniques.

Bridging the Gap Between Artificial Neural Networks and Kernel Regressions for Vector-Valued Problems in Microwave Applications

Thanks to their convex formulation, kernel regressions have shown an improved accuracy with respect to artificial neural network (ANN) structures in regression problems where a reduced set of training samples are available. However, despite the above interesting features, kernel regressions are inherently less flexible than ANN structures since their implementations are usually limited to scalar-output regression problems. This article presents a vector-valued (multioutput) formulation of the kernel ridge regression (KRR) aimed at bridging the gap between multioutput ANN structures and scalar kernel-based approaches. The proposed vector-valued KRR relies on a generalized definition of the reproducing kernel Hilbert space (RKHS) and on a new multioutput kernel structure. The mathematical background of the proposed vector-valued formulation is extensively discussed together with different matrix kernel functions and training schemes. Moreover, a compression strategy based on the Nystrom approximation is presented to reduce the computational complexity of the model training. The effectiveness and the performance of the proposed vector-valued KRR are discussed on an illustrative example consisting of a high-speed link and on the optimization of a Doherty amplifier.

Using Artificial Neural Networks for Predicting Ship Fuel Consumption

Abstract In marine vessel operations, fuel costs are major operating costs which affect the overall profitability of the maritime transport industry. The effective enhancement of using ship fuel will increase ship operation efficiency. Since ship fuel consumption depends on different factors, such as weather, cruising condition, cargo load, and engine condition, it is difficult to assess the fuel consumption pattern for various types of ships. Most traditional statistical methods do not consider these factors when predicting marine vessel fuel consumption. With technological development, different statistical models have been developed for estimating fuel consumption patterns based on ship data. Artificial Neural Networks (ANN) are some of the most effective artificial methods for modelling and validating marine vessel fuel consumption. The application of ANN in maritime transport improves the accuracy of the regression models developed for analysing interactive relationships between various factors. The present review sheds light on consolidating the works carried out in predicting ship fuel consumption using ANN, with an emphasis on topics such as ANN structure, application and prediction algorithms. Future research directions are also proposed and the present review can be a benchmark for mathematical modelling of ship fuel consumption using ANN.

Neural Network Algorithm With Reinforcement Learning for Parameters Extraction of Photovoltaic Models.

This research focuses on the application of artificial neural networks (ANNs) on parameters extraction of photovoltaic (PV) models. Extracting parameters of the PV models accurately is crucial to control and optimize PV systems. Although many algorithms have been proposed to address this issue, how to extract the parameters of the PV models accurately and reliably is still a great challenge. Neural network algorithm (NNA) is a recently reported metaheuristic algorithm. NNA is inspired by ANNs. Benefiting from the unique structure of ANNs, NNA shows excellent global search ability. However, NNA faces the challenge of slow convergence rate and local optima stagnation in solving complex optimization problems. This article presents an improved NNA, named neural network algorithm with reinforcement learning (RLNNA), for extracting parameters of the PV models. In RLNNA, three strategies, namely modification factor with reinforcement learning (RL), transfer operator with historical population, and feedback operator, are designed to overcome the challenge of NNA. To verify the performance of RLNNA, it is employed to extract the parameters of the three PV models. Experimental results show that RLNNA can extract the parameters of the considered PV models with higher accuracy and stronger stability compared with NNA and the other 12 powerful algorithms, which fully indicates the effectiveness of the improved strategies.

Insight into nettle straw pyrolysis: Multicomponent kinetics, gas emissions and machine learning models

This study aimed to comprehensively investigate the pyrolysis process of nettle straw by analyzing its multicomponent kinetics modeling, kinetics and thermodynamic parameters, pyrolysis performance, reaction mechanisms, gas emissions, correlation and statistical analysis, as well as feasibility for industrial application. The average apparent activation energies for pseudo-hemicellulose (PS-HEC), cellulose (PS-CEL) and lignin (PS-LIG) were estimated as 147.41, 182.66, and 197.12 kJ/mol, respectively, using four model-free methods: Ozawa-Flynn-Wall, Kissinger-Akahira-Sunose, Starink, and Tang. The pyrolysis reaction mechanisms were best described by diffusional and order-based models, with reaction mechanism functions of f(α) = [3/2(1-α)2/3]/[1-(1-α)1/3], f(α) = (1-α)1.5, and f(α) = (1-α)3 for PS-HEC, PS-CEL, and PS-LIG, respectively. TGA-FTIR analysis revealed that the gas emissions during nettle straw pyrolysis followed the order of CO > CO2 > C-O(H) > H2O > SO2 > CO > CH4 > NH3 > HCN. Best learning rate and artificial neural network (ANN) structure model were optimized. The optimal learning rate and ANN model structure were 0.0013 and ANN (5 *9 *1), respectively. NS has the potential to serve as a viable alternative to biofuel due to its certain energy, low price, abundant reserves, good pyrolysis performance, high volatile content, and low nitrogen and sulfur levels.

An investigation of artificial neural network structure and its effects on the estimation of the low‐cycle fatigue parameters of various steels

AbstractArtificial neural networks (ANNs) are a widely used machine learning approach for estimating low‐cycle fatigue parameters. ANN structure has its parameters such as hidden layers, hidden neurons, activation functions, training functions, and so forth, and these parameters have a significant influence over the results. Three hidden layer combinations, the hidden neurons ranging from 1 to 25, and different activation functions like hyperbolic tangent sigmoid (tansig), logistic sigmoid (logsig), and linear (purelin) were used, and their effects on the low‐cycle fatigue parameter estimation were investigated to determine optimal ANN structure. Based on the results, suggestions regarding ANN structure for the estimation of the low‐cycle fatigue parameters and transition fatigue life were presented. For the output layer and hidden layers, the most suitable activation function was tansig. The optimal hidden neuron range has been found between 4 and 9. The neural network structure with one hidden layer was determined to be most suitable in terms of less knowledge, structural complexity, and computational time and power.

Fractional-order ascent maximum mixture correntropy criterion for FLANNs based multi-channel nonlinear active noise control

The filtered-s maximum correntropy criterion (FsMCC) algorithm was developed for nonlinear active noise control (NANC) in impulsive noise environments with the use of functional linked artificial neural networks (FLANNs) structure. However, the kernel-width of the FsMCC is constant, which is dull and cannot self-regulate to achieve both the fast convergence and low steady-state error performance. For better convergence ability, in this paper, a fractional-order gradient ascent filtered-s maximum correntropy criterion algorithm (FGA-FsMCC) is presented for multi-channel NANC system. The fractional lower-order moments supersede the non-existent second-order moments. To come up with a more reliable algorithm, by merging two kernels with complementary width, the fractional-order gradient ascent filtered-s mixture correntropy criterion (FGA-FsMMCC) was presented. The computational complexity and convergence conditions are analyzed. To verify the robustness of proposed algorithms, simulations are carried out to discuss parameters and control effects of algorithms for alpha-stable noise in different nonlinear environments.

Modeling the viscosity of ionic liquids using combined friction theory with perturbed hard-chain equation of state and neural network approaches

Fast and accurate prediction and correlation of thermophysical properties are always important concerns for researchers and engineers. This work is the extension of our earlier friction theory (FT)-based approach for estimating the dynamic viscosities of pure ionic liquids using a perturbed hard-chain equation of state. The model used three molecular parameters along with 5 correlation coefficients in the friction terms. The correlation coefficients were obtained using 167 data points at 0.1 MPa with an average absolute deviation (AAD) of 0.61 % and then used to predict the viscosity values at higher pressures. Using the FT-based model, 2907 experimental viscosity data for 20 pure ILs were predicted. The calculations were carried out in 273–373 K range and pressures up to 298.9 MPa for 2907 data points with an AAD of 2.63 %. The present FT-based model was not able to accurately predict the viscosities of ILs outside the temperature range used for the fit procedure. Also, comparing the precision of our FT-based model with two rough-hard-sphere theory-based models of Hosseini et al. and Hossain-Teja as well as the group-contribution method led to AADs of 2.46 % (for 2664 data points), 2.65 % (for 1605 data points), 3.25 % (for 890 data points), and 7.73 % (for 2558 data points, respectively. Viscosities of several mixtures of IL + IL type were also correlated using the present model along with some simple combining equations and a binary interaction parameter as well. The AAD of the calculated viscosities from the experimental ones was found to be 5.10 % for 960 data points examined. In addition, a neural network was trained on 3074 data points to represent the viscosity of studied pure ILs. The artificial neural network (ANN) structure utilized two layers, four input parameters, and 13 neurons for each layer. This specific structure of ANN was carefully designed to ensure an optimal performance in predicting the viscosity of ILs. The overall AAD was obtained 1.44 %.

Specific Attenuation Estimation in Perturbed Radio Wave Propagation Using Artificial Neural Networks

Path loss modeling is a crucial consideration in radio engineering for wireless networks. Over the years, diverse techniques have been implemented in attempts to accurately predict path loss across a given terrain. In this study, path loss predictors created on the bases of artificial neural networks (ANN) were used to estimate path loss across a rural section of the Nigerian middle-belt grassland. The ANN structures considered were the Generalized Regression Neural network (GRNN) and the Radial Basis Function Neural Network (RBFNN), which exhibit a few differences and similarities. These ANN based predictors were trained, validated and tested for path loss prediction using path loss values computed from received power measured at 900MHz from six Base Transceiver Stations (BTSs) situated along the rural terrain. Findings show that the RBFNN predictor with a Root Mean Squared Error (RMSE) of 5.17dB and the GRNN with 4.9dB are slightly more accurate than the COST 231 Hata model with 6.64dB, while the Hata-Okumura with 25.78dB is simply not suitable for the terrain under investigation. Overall, the GRNN, which proffers a 26.21% improvement over the COST 231 Hata is recommended for the terrain in question.

Energy-Efficient Hardware Implementation of Fully Connected Artificial Neural Networks Using Approximate Arithmetic Blocks.

In this paper, we explore efficient hardware implementation of feedforward artificial neural networks (ANNs) using approximate adders and multipliers. Due to a large area requirement in a parallel architecture, the ANNs are implemented under the time-multiplexed architecture where computing resources are re-used in the multiply accumulate (MAC) blocks. The efficient hardware implementation of ANNs is realized by replacing the exact adders and multipliers in the MAC blocks by the approximate ones taking into account the hardware accuracy. Additionally, an algorithm to determine the approximate level of multipliers and adders due to the expected accuracy is proposed. As an application, the MNIST and SVHN databases are considered. To examine the efficiency of the proposed method, various architectures and structures of ANNs are realized. Experimental results show that the ANNs designed using the proposed approximate multiplier have a smaller area and consume less energy than those designed using previously proposed prominent approximate multipliers. It is also observed that the use of both approximate adders and multipliers yields, respectively, up to 50% and 10% reduction in energy consumption and area of the ANN design with a small deviation or better hardware accuracy when compared to the exact adders and multipliers.

Cardiovascular diseases prediction by machine learning incorporation with deep learning.

It is yet unknown what causes cardiovascular disease (CVD), but we do know that it is associated with a high risk of death, as well as severe morbidity and disability. There is an urgent need for AI-based technologies that are able to promptly and reliably predict the future outcomes of individuals who have cardiovascular disease. The Internet of Things (IoT) is serving as a driving force behind the development of CVD prediction. In order to analyse and make predictions based on the data that IoT devices receive, machine learning (ML) is used. Traditional machine learning algorithms are unable to take differences in the data into account and have a low level of accuracy in their model predictions. This research presents a collection of machine learning models that can be used to address this problem. These models take into account the data observation mechanisms and training procedures of a number of different algorithms. In order to verify the efficacy of our strategy, we combined the Heart Dataset with other classification models. The proposed method provides nearly 96 percent of accuracy result than other existing methods and the complete analysis over several metrics has been analysed and provided. Research in the field of deep learning will benefit from additional data from a large number of medical institutions, which may be used for the development of artificial neural network structures.

Two-Module Neural Network Method of Information Processing in Gas Analysis Systems

Purpose of research: reduction of additional errors in measuring gas concentrations in gas analytical systems (GS) caused by the sensitivity of semiconductor sensors to non-target components of gas mixtures, ambient temperature and humidity. To develop and test a two-module neural network method for processing information in a GS, which allows automating the processes of generating training data and searching for the optimal structure of artificial neural networks (ANNs), reducing errors in reproducing the characteristics of sensors by replacing their mathematical models with neural networks.Methods. Theory of artificial neural networks, numerical methods, simulation methods. To evaluate the effectiveness of the proposed solution, the relative error (d), standard deviation (RMS) were calculated, and comparison with analogues was carried out.Results: a two-module neural network method for processing information in a GS has been studied. Numerical modeling was used to carry out experimental studies on the choice of optimal ANN structures, the volume and composition of training data. In the course of experimental studies, the errors of generating training data using ANN (less than 5%) and determining the concentrations of detected gases under conditions of fluctuations in the parameters of the air environment and the composition of the gas mixture (less than 4%) were calculated.Conclusion. A two-module neural network method for information processing is proposed, which is distinguished by the use of two successive modules of multilayer neural networks for generating training data and processing information coming from the GS sensor unit. The use of an auxiliary module makes it possible to compress the initial data, unify and automate the process of their generation, as well as improve the accuracy of reproduction of multiparameter sensor conversion functions, in comparison with alternative methods. Results of experimental studies of the effectiveness of using the information processing method to reduce additional errors in the quantitative determination of the composition of the air environment under conditions of parameter fluctuations are presented.

Efficient training of two ANNs using four meta-heuristic algorithms for predicting the FRP strength

In recent years, artificial neural network (ANN) is one of the popular and effective machine learning models that can be used to accurately predict fiber reinforced polymer (FRP) strength. However, the ANN structure and parameters are usually are chosen by experience. In this study, the aim is to use the combination of meta-heuristic algorithm with two different types of artificial neural network structure to optimize the parameters of the feed forward backpropagation and radial basis function networks. In this paper, Particle swarm optimization (PSO), Genetic algorithm (GA), Colliding bodies optimization (CBO), Enhanced colliding bodies optimization (ECBO) algorithms are used to combine with ANNs.A total of 223 test data on Carbon FRP (CFRP) collected from the available literature were used to generate training and test data sets. Various validation criteria such as mean square error, root mean square error and correlation coefficient (R) are used to validate the models. These models consider the effects of concrete compressive strength, concrete sample diameter, concrete sample length, fiber elastic modulus, fiber thickness, fiber strength on the ultimate strength of FRP-concrete.It is shown that ECBO algorithm provide better results and higher accuracy. Also, the comparison of the results of using two neural networks, feed forward backpropagation (FFB) and radial basis function (RBF), indicates that the lower error percentage and more accurate results are related to the FFB neural networks combined with ECBO.

Design of a novel robust adaptive neural network‐based fractional‐order proportional‐integrated‐derivative controller on DC/DC Boost converter

AbstractDC/ DC Boost converter has a right half‐plane zero structure called a non‐minimum phase system, which presents several challenging constraints for designing well‐behaved control techniques. The Fractional‐Order concept as a beneficial scheme provides several advantages, such as lower sensitivity to noise and parametric variation. For this purpose, a Fractional‐order Proportional‐Integrated‐Derivative (FOPID) controller is designed for the Boost converter. On the other hand, for wider ranges of disturbances, including parametric variations, load uncertainty, supply voltage variation, and noise, this approach shows an unsuitable practical application based on its fixed gain values; therefore, the control parameters need to be optimized again to provide ideal operations. An Artificial Neural Network structure (ANNs) is adopted here to optimize the gains of the FOPID in challenging conditions. This method suffers from higher complexity and slower dynamics in practical applications. A single‐layer ANNs is designed to reduce the complexity; also, particle swarm optimization (PSO) algorithm is utilized in this decision‐making structure to provide better results with its online mechanism. Furthermore, the Black‐box technique is applied for the proposed system, which does not require an accurate mathematical model resulting in a lower computational burden, easy implementation, and lower dependency on the states of the model. The Artificial Neural Network structure can optimize the FOPID gains real‐time, even in severe challenging conditions, efficiently. To better examine the superiority of the proposed method, conventional Proportional‐Integrated‐Derivative (PID) and the FOPID controllers are proposed to drive a comparison with this work, which are tuned by the PSO optimization algorithm. The evaluation of simulation results demonstrates that the proposed control scheme is suitable not only for preserving stability but also for compensating for the disturbances and uncertainties in the Boost converter, properly.