Feed-forward Back-propagation Artificial Neural Network Research Articles

Artificial neural network based delamination prediction in composite plates using vibration signals

Dynamic loading on composite components may induce damages such as cracks, delaminations, etc. and development of an early damage detection technique for delaminations is one of the most important aspects in ensuring the integrity and safety of composite components. The presence of damages such as delaminations on the composites reduces its stiffness and further changes the dynamic behaviour of the structures. As the loss in stiffness leads to changes in the natural frequencies, mode shapes, and other aspects of the structure, vibration analysis may be the ideal technique to employ in this case. In this research work, the supervised feed-forward multilayer back-propagation Artificial Neural Network (ANN) is used to determine the position and area of delaminations in GFRP plates using changes in natural frequencies as inputs. The natural frequencies were obtained by finite element analysis and results are validated by experimentation. The findings show that the suggested technique can satisfactorily estimate the location and extent of delaminations in composite plates.

Artificial Neural Network Model for the Prediction of Thermal Conductivity of Saturated Liquid Refrigerants and n-Alkanes.

In this paper, a feed-forward back-propagation artificial neural network (ANN) is proposed to correlate and predict the thermal conductivity from the triple point temperature up to 0.98 times critical temperature (T c) for 23 refrigerants and 11 n-alkanes. It requires the temperature (T) as well as the molecular mass (M), acentric factor (ω), critical temperature, and critical pressure (P c) as input variables. The optimal ANN model is obtained by a trial-and-error procedure and consists of the input layer and the output layer together with one hidden layer with seven neurons. This ANN model can not only correlate the thermal conductivity but also accurately predict the thermal conductivity of refrigerants and n-alkanes. The correlation coefficients (R) in the training and testing phases are 0.9994 and 0.9993, respectively. Furthermore, the average absolute deviation (AAD) values are less than 1% for 14 out of 34 fluids, less than 2% for 28 fluids, and less than 4.5% for all the considered fluids.

Application of Artificial Neural Network to Predict the Properties of Permeable Concrete

The structure of permeable concrete has been the primary reason for its use in construction. Permeable concrete is composed of water, cement, aggregate, and little- to no-fines resulting in the presence of a significant number of voids. This makes permeable concrete an ideal solution to water accumulation issues as it acts as a drainage system. This study employs a feedforward backpropagation artificial neural network model that combines experimental laboratory data from previous studies with appropriate network architectures and training techniques. The purpose of the analysis is to develop a reliable functional relationship, based on water-cement ratio, aggregate-cement ratio, and density parameters, with which to estimate the compressive strength, porosity, and water permeability of permeable concrete. Multiple linear regression correlations are also established to predict and correlate these inputs and outputs. The two derived methods are then compared and discussed. The results reveal that ANN is better to anticipate the permeable concrete properties than regression analysis.

Prediction of gamma ray spectrum for 22Na source by feed forward back propagation ANN model

The radiation has been used in a variety of different fields since its discovery and thus its measurement becomes vital in these industries. Different type detector may be used to measure gamma rays depends on the purposes of measurements. Gamma ray energy spectrum is an important to determine either elemental analysing of a sample or radiation shielding purposes. On the other hand, Artificial Neural Network (ANN) may be used to predict and analysing of gamma-ray spectrum. In this study, gamma ray spectrum from 22Na source detected in NaI (Tl) detector was estimated by ANN. There have been installed ten different ANN models to find the network structure that produces the best predictive value for the gamma ray spectrum NaI (Tl) Detector. Estimation study has been continued with the ANN model with be possessed of lowest error value. ANN model was created by using energy, distance and gamma-rays energy spectrum (called Io) values. In the ANN model developed using the feed forward back propagation algorithm, were used artificial neurons two in the input layer, ten in the hidden layer and one in the output layer. For the case of present work, the experimental data was used 70% for education, 20% for validation and 10% for testing. The estimated values obtained with the ANN model were compared with the experimental results and a good correlation has been found between them (R2 = 0.99).

Peak ground acceleration prediction using artificial neural networks for Kachchh, Gujarat, India

The uncertainty in the empirical ground motion prediction models (GMMs) for any region depends on several parameters. In the present work, we apply an artificial neural network (ANN) to design a GMM of peak ground acceleration (PGA) for Kachchh, Gujarat, India, utilizing independent input parameters viz., moment magnitudes, hypocentral distances, focal depths and site proxy (in terms of average seismic shear-wave velocity from the surface to a depth of 30 m (Vs30)). The study has been performed using a PGA dataset consisting of eight engineering seismoscope records of the 2001 Mw7.7 Bhuj earthquake and 237 strong-motion records of 32 significant Bhuj aftershocks of Mw3.3–5.6 (during 2002–2008) with epicentral distances ranging from 1.0 to 288 km. We apply a feed-forward back propagation ANN method with 8 hidden nodes, which is found to be optimal for the selected PGA database and input–output mapping. The standard deviation of the error has been utilized to examine the performance of our model. We also test the ground motion predictability of our ANN model using real recordings of the 2001 Bhuj mainshock, two Mw5.6 Kachchh aftershocks and the 1999 Mw6.4 Chamoli mainshock. The standard deviation of PGA prediction error estimates in log10 units is found to be ± 0.2554. Also, the model predictability of our ANN model suggests a good prediction of the PGA for earthquakes of Mw5.6–7.7, which are occurring in Kachchh, Gujarat, India.

Development of artificial intelligence‐based neural network prediction model for responses of additive manufactured polylactic acid parts

AbstractFused deposition modeling (FDM) is one of the most economical and popular technology amongst numerous additive manufacturing techniques. The quality of FDM fabricated parts is highly sensitive to the production parameters. Thus, in the present work, an investigation on the FDM printed polylactic acid parts has been performed considering six printing process parameters, that is, nozzle diameter, build orientation, raster pattern, layer height and print speed to develop the feedforward backpropagation (FFBP) artificial neural network prediction model for the prediction of responses, namely, tensile strength, material consumption, build time and surface quality. Tensile specimens as per L27 orthogonal array are printed considering the various combination of parameters. The printed samples have been subjected to tensile strength testing, surface roughness measurement, build time recording, and material consumption evaluation. The highest tensile strength of 57.633 MPa, lowest surface roughness of 1.71 μm, lowest build time of 0.35 h and lowest material consumption of 7.8 g are observed. The experimental results have been used to develop the artificial intelligence‐based prediction model through FFBP algorithm and sigmoid transfer function to predict the responses. The best performance of the developed neural network with R2 for testing (0.99343), training (0.99366), and validation (0.99372) of data is recorded for prediction of responses with minimum percentage error. The study concluded that developed model is capable of predicting the responses of FDM process according to the input process parameters.

Prediction of the Degree of Polymerization in Transformer Cellulose Insulation Using the Feedforward Backpropagation Artificial Neural Network

The life expectancy of power transformers is primarily determined by the integrity of the insulating oil and cellulose paper between the conductor turns, phases and phase to earth. During the course of their in-service lifetime, the solid insulating system of windings is contingent on long-standing ageing and decomposition. The decomposition of the cellulose paper insulation is strikingly grievous, as it reduces the tensile strength of the cellulose paper and can trigger premature failure. The latter can trigger premature failure, and to realize at which point during the operational life this may occur is a daunting task. Various methods of estimating the DP have been proposed in the literature; however, these methods yield different results, making it difficult to accurately estimate a reliable DP. In this work, a novel approach based on the Feedforward Backpropagation Artificial Neural Network has been proposed to predict the amount of DP in transformer cellulose insulation. Presently, no ANN model has been proposed to predict the remaining DP using 2FAL concentration. A databank comprising 100 data sets—70 for training and 30 for testing—is used to develop the proposed ANN using 2-furaldehyde (2FAL) as an input and DP as an output. The proposed model yields a correlation coefficient of 0.958 for training, 0.915 for validation, 0.996 for testing and an overall correlation of 0.958 for the model.

Artificial Bee and Ant Colony-assisted Performance Improvements in Artificial Neural Network-based Rotor Fault Detection

Asynchronous motors are the most commonly used types of motor in the industry. They are preferred because of their ease of control and reasonable cost. Since it is not desirable to suspend production in factories, it is required that motor failures used in production lines be detected quickly and easily. In this article, sound signals were recorded during the operation of the asynchronous motor, which is operational and with a rotor bar crack; and filtering, normalization, and Fast Fourier Transform were performed. The detection of rotor broken bar error was examined using the feed-forward backpropagation Artificial Neural Network (ANN) method. With intuitive algorithms such as the artificial bee colony and artificial ant colony, improvements to the ANN results were investigated. The experimental results verified that intuitive algorithms can improve the estimation performance of the neural network.

Feedforward backpropagation artificial neural networks for predicting mechanical responses in complex nonlinear structures: A study on a long bone

Feedforward backpropagation artificial neural networks (ANNs) have been increasingly employed in many engineering practices concerning materials modeling. Despite their extensive applications, how to achieve successfully trained ANNs is not thoroughly explained in the literature, nor are there lucid discussions to delineate influential parameters obtained from analyses. Long bones are composite materials possessing nonhomogeneous and anisotropic properties, and their mechanical responses exhibit dependency on numerous variables. Material complexity hinders researchers from arriving at a consensus in implementing an optimal constitutive model or encourages them to adopt a simple constitutive model including many simplifying assumptions. However, such exceptional features and engineering challenges make long bones materials worth investigating, enriching our comprehension of complex engineering structures using novel techniques where traditional methods may present limitations. This paper reports on the prediction of loading, displacement, load and displacement simultaneously, and strains using feedforward backpropagation ANNs trained with experimental recordings. The technique was used to find optimum network structures (architectures) that encompass the best prediction ability. To enhance predictions, the influence of several elements such as a network training algorithm, injecting noise to datasets prior to training, the level of injected noise which directly affects model fitting and regularization, and data normalization prior to training were investigated and discussed. Essential parameters influencing decision making in identifying well-trained and well-generalized ANNs were elaborated. A considerable emphasis in this study was placed on examining the generalization ability of the already trained and tested ANNs, thus guaranteeing unbiased models that avoided overfitting. Gaining favorable outcomes in this study required three years of performing experiments and data collection before establishing the networks. The subsequent training, testing, and determination of the generalization of more than 60,000 ANNs are promising and will assist researchers in comprehending mechanical responses of complicated engineering structures that exhibit peculiar nonlinear properties.

Implementation of cognitive radio networks for optimum spectrum utilization through feed forward backpropagation artificial neural network

In wireless Radio Frequency (RF) networks, a huge amount of available bandwidth of the RF spectrum is wasted, because the users are not able to fully utilize the frequency band of the spectrum. The Cognitive Radio Networks provide an effective solution for optimum utilization of the frequency band of the RF spectrum. Cognitive Radio Networks consist of two types of users: primary users, which utilizes their allocated license frequency band of the RF spectrum and secondary users, which uses the license frequency band of the primary users when they are not using the RF spectrum. Primary users pre-empt the secondary users when they want to use the RF spectrum. To know when the spread spectrum is not occupied by the primary users, the secondary users must have some mechanism to sense the environment, learn the parameters and take a decision on the basis of it. The users of Cognitive Radio Networks enhance these capabilities by using Artificial Neural Networks (ANN). This paper discusses how the presence of primary users in the particular frequency band of the spectrum could be detected by the secondary users by using Feed Forward Back Propagation ANN. For this a CRN simulation model is designed and developed which uses a Feed forward back propagation ANN for optimum use of the spectrum.

Soil permeability enhancement using pneumatic fracturing coupled by vacuum extraction for in-situ remediation: Pilot-scale tests with an artificial neural network model

Enhancing soil permeability is of huge practical significance for the in-situ chemical oxidation in soil and groundwater remediation. Hence, we conducted pilot-scale (2 m3 of soil) pneumatic fracturing (PF) and pneumatic fracturing-vacuum extraction (PFV) experiments for improving the fluid infiltration rate in low permeable soil zone in this study. Moreover, the correlation interpretation for soil permeability enhancement suffers from limited databases and complexities (i.e., geological properties) due to many factors that affect the fluid injection performance toward low permeability soil. Further, this is the first paper to introduce a novel approach using a soft computational model, a feedforward backpropagation artificial neural network (FFBP-ANN), to predict the infiltration coefficient for PF or PFV treated soils. As a result, the PV and PFV methods significantly enhanced the infiltration coefficients. Notably, the established FFBP-ANN model with the configuration of eight neurons associated with one hidden layer connected by tangent sigmoid transfer function and trained by Levenberg-Marquart backpropagation algorithm achieved the 0.999 of regression with 0.001 of mean square error accuracy performance. Therefore, this study shows that PFV can significantly enhance the infiltration coefficient, and the computational FFBP-ANN models can help extend the infiltration coefficient estimation for low permeable soil.

Assessment and development of air quality forecasting models by the combination of PCA and ANN at metropolis Kolkata, India

PM10, PM2.5, SO2, NO2, Pb, As and Ni were monitored at Jadavpur, Kolkata over the time span of October 2016 to March 2017. The air quality of the study area was assessed in terms of air quality index (AQI) which is calculated on the assessed concentration of the pollutants. The result of the study showed significant seasonal variance and the ambient air of the study area is highly polluted in terms of particulate matters, NO2 and metallic pollutants. AQI study reveals that PM2.5 was found responsible for this degraded air quality of Kolkata. Principal component analysis (PCA) was used to identify possible contributors i.e. i) construction work, windblown dust, road dust and earth crust (variance- 28.98%); ii) automobile exhaust (variance- 20.47%); iii) brake wear source, road dust and tire wear (variance- 18.07%) of air pollution in the study area. Multiple linear regression (MLR) and feedforward backpropagation artificial neural network (FFBPANN) were used in combination with PCA to predict the future PM2.5 concentrations. Several performance indexes were used to assess the degree of accuracy and the magnitude of error in the model predicted values. Overall performance index establishes the fact that PCA-FFBP-ANN model most accurately predicts the future PM2.5 concentrations.

ANN-based structure-viscosity relationship model of multicomponent slags for production design in mineral wool

The production of mineral wool from waste high-temperature slags is considered eco-friendly and economical. However, this application is facing a great challenge from an unstable production process caused by the variable compositions and properties of slags. Especially, the viscosity of slags, as a significant property, governs the processes of melt homogenization, tapping, and fiber formation. Here, we designed a structure-property relationship model to describe the development of slag viscosity over a broad range of composition and temperature. A feed-forward back-propagation artificial neural network was adopted as the basis to establish quantitative relationships between effective variables and viscosity. A comprehensive database containing 5474 reliable laboratory-based samples was constructed and further developed using the micro-scale structural features. The importance of structural features was evaluated using the statistical connection weight approach and 74 kinds of first-nearest-neighbor pairs and 4 medium-range characteristics (three kinds of oxygen species and the degree of depolymerization of network) were selected and integrated into the model. The results indicated the obtained model exhibited excellent performance, with a testing mean square error of 0.0592 and a testing Pearson’s correlation coefficient of 0.9966. The model was compared with several other common viscosity models and also showed the optimal predictive ability. Furthermore, the model was successfully applied in the production of mineral wool from three complex slags and the high-quality mineral wool product was achieved.

Modelling tropospheric ozone variations using artificial neural networks: A case study on the Black Sea coast (Russian Federation)

This study focuses on modelling tropospheric ozone using neural networks to predict its concentration depending on environmental parameters. Predicting tropospheric ozone concentration is an important task, especially in recreational areas, since it harms human health. Considering the great complexity in describing the mechanisms of formation and destruction of tropospheric ozone and their dependence on a number of factors (temperature, humidity, pressure, wind speed and direction) the authors have suggested using Neural Networks to predict it. The selection of the network configuration and the algorithm for its training largely depends on the type of initial data available to researchers. The following fundamental factors were taken into account: temperature, humidity, and wind direction. The selected Neural Networks have been applied to a dataset obtained from the Russian Federation. From the obtained results, the best forecasting accuracy was achieved by using a Feed-Forward Back-Propagation Artificial Neural Network with 3 layers. The accuracy of prediction of the artificial neural networks against the measured data was evaluated using the Index of Agreement (IOA), which was estimated at 0.87, which equals other work in this field. This level of accuracy is equivalent to previous advances, but the standard software and built-in neural network configurations were used. The presented results have also confirmed that the wind speed and direction have a significant impact on the forecast accuracy – excluding wind speed reduces the IOA to 0.839; excluding wind speed and direction reduces the IOA to 0.0.807. • Modelling data on the content of tropospheric ozone in the Karadag landscape protection area. • Artificial neural network Feed Forward Back-propagation performs better than others. • Seven artificial neural networks for tropospheric ozone prediction are presented. • A significant influence of wind parameters on forecast accuracy tropospheric ozone concentration. • The accuracy of prediction of the artificial neural networks was estimated at 87%, which equals other work in this field.

Pull-in Phenomenon in the Electrostatically Micro-switch Suspended between Two Conductive Plates using the Artificial Neural Network

Artificial Neural Networks (ANN) are designed to evaluate the pull-in voltage of MEMS switches. The mathematical model of a micro-switch subjected to electrostatic force is preliminarily illustrated to get the relevant equations providing static deflection and pull-in voltage. Adopting the Step-by-Step Linearization Method together with a Galerkin-based reduced order model, numerical results in terms of pull-in voltage are obtained to be employed in the training process of ANN. Then, feed forward back propagation ANNs are designed and a learning process based on the Levenberg-Marquardt method is performed. The ability of designed neural networks to determine pull-in voltage have been compared with previous results presented in experimental and theoretical studies and it has been shown that the presented method has a good ability to approximate the threshold voltage of micro switch. Furthermore, the geometric and physical effect of the micro-switch on the pull-in voltage was also examined using these designed networks and relevant findings were provided.

Vibro-Fluidized Bed Drying of Pumpkin Seeds: Assessment of Mathematical and Artificial Neural Network Models for Drying Kinetics

Pumpkin seeds were dried in a vibro-fluidized bed dryer (VFBD) at different temperatures at optimized vibration intensity of 4.26 and 4 m/s air velocity. The drying characteristics were mapped employing semiempirical models and Artificial Neural Network (ANN). Prediction of drying behavior of pumpkin seeds was done using semiempirical models, of which, one was preferred as it indicated the best statistical indicators. Two-term model showed the best fit of data with R2 − 0.999, and lowest χ2 − 1.03 × 10−4 and MSE 7.55 × 10−5. A feedforward backpropagation ANN model was trained by the Levenberg–Marquardt training algorithm using a TANSIGMOID activation function with 2-10-2 topology. Performance assessment of ANN showed better prediction of drying behavior with R2 = 0.9967 and MSE = 5.21 × 10−5 for moisture content, and R2 = 0.9963 and MSE = 2.42 × 10−5 for moisture ratio than mathematical models. In general, the prediction of drying kinetics and other drying parameters was more precise in the ANN technique as compared to semiempirical models. The diffusion coefficient, Biot number, and hm increased from 1.12 × 10−9 ± 3.62 × 10−10 to 1.98 × 10−9 ± 4.61 × 10−10 m2/s, 0.51 ± 0.01 to 0.60 ± 0.01, and 1.49 × 10−7 ± 4.89 × 10−8 to 3.10 × 10−7 ± 7.13 × 10−8 m/s, respectively, as temperature elevated from 40 to 60°C. Arrhenius’s equation was used to the obtain the activation energy of 32.71 ± 1.05 kJ/mol.

Machine learning approaches for predicting geometric and mechanical characteristics for single P420 laser beads clad onto an AISI 1018 substrate

The final mechanical and physical properties should be predicted in tandem with the bead geometry characteristics for effective additive manufacturing (AM) solutions for processes such as directed energy deposition. Experimental approaches to investigate the final geometry and the mechanical properties are costly, and simulation solutions are time-consuming. Alternative artificial intelligent (AI) systems are explored as they are a powerful approach to predict such properties. In the present study, the geometrical properties as well as the mechanical properties (residual stress and hardness) for single bead clads are investigated. Experimental data is used to calibrate multi-physics finite element models, and both data sets are used to seed the AI models. The adaptive neuro-fuzzy inference system (ANFIS) and a feed-forward back-propagation artificial neural network (ANN) system are utilized to explore their effectiveness in the 1D (discrete values), 2D (bead cross-sections), and 3D (complete bead) domains. The prediction results are evaluated using the mean relative error measure. The ANFIS predictions are more precise than those from the ANN for the 1D and 2D domains, but the ANN had less error for the 3D scenario. These models are capable of predicting the geometrical and the mechanical properties values very well, including capturing the mechanical properties in transient regions; however, this research should be extended for multi-bead scenarios before a conclusive “best approach” strategy can be determined.

İnsansız Hava Araçlarının (İHA) Şifrelenmiş Wi-Fi Trafiği Üzerinden Çalışma Modlarının Yapay Sinir Ağları Aracılığı İle Belirlenmesi

Developing technology has also made the Unmanned Aerial Vehicles (UAV) widespread. While UAVs provide beneficial use in many sectors from engineering solutions to visual arts, they also come up with malicious uses and can even be used as a tool for committing crimes. Although the states are trying to register its use with legislation in order to prevent this problem, the problem has not been completely eliminated. The most important problem we face about UAVs is to be able to percept quickly and effectively for what purpose they are flying over a certain region. Although previous studies in the literature were partially successful in solving this problem, it could not be considered as an effective solution due to high costs and long detection time. In this study, the encrypted wi-fi traffic was tried to be defined by the data packet size analysis method to determine the operating modes of the UAVs. Since the amount of data and data processing speed are the most important factors in the detection of UAVs, processes based on artificial intelligence and machine learning have been applied. Using the feed-forward backpropagation artificial neural network method, the operating modes of the UAVs were determined and a success rate of 99.29% was achieved.

Use of artificial neural network and adaptive neuro-fuzzy inference system for prediction of biogas production from spearmint essential oil wastewater treatment in up-flow anaerobic sludge blanket reactor

Two artificial intelligence (AI)-based models, namely feed-forward backpropagation artificial neural network (ANN) and multi-layer adaptive neuro-fuzzy inference system (ANFIS), were developed to estimate the biogas production in an up-flow anaerobic sludge blanket (UASB) reactor. The models' input variables including influent chemical oxygen demand (COD), pH, effluent mixed liquor suspended solids, effluent mixed liquor volatile suspended solids, turbidity removal, oil and grease removal, COD removal, phenol removal, and effluent volatile fatty acids and alkalinity were collected from an UASB reactor fed with spearmint essential oil wastewater (SEOW) during 141 consecutive operating days. The determination coefficient, root mean square error, and relative root mean the ANN model's square error reached 0.975, 2650 mL/d, and 0.234%, respectively, while those of ANFIS model reached 0.956, 3517 mL/d, and 0.315%, respectively. The results achieved herein demonstrated that two AI-based models were successful to estimate the biogas production in a lab-scale UASB reactor treating SEOW with high accuracy and low error.

Estimation of unsteady hydromagnetic Williamson fluid flow in a radiative surface through numerical and artificial neural network modeling

In current investigation, a novel implementation of intelligent numerical computing solver based on multi-layer perceptron (MLP) feed-forward back-propagation artificial neural networks (ANN) with the Levenberg–Marquard algorithm is provided to interpret heat generation/absorption and radiation phenomenon in unsteady electrically conducting Williamson liquid flow along porous stretching surface. Heat phenomenon is investigated by taking convective boundary condition along with both velocity and thermal slip phenomena. The original nonlinear coupled PDEs representing the fluidic model are transformed to an analogous nonlinear ODEs system via incorporating appropriate transformations. A data set for proposed MLP-ANN is generated for various scenarios of fluidic model by variation of involved pertinent parameters via Galerkin weighted residual method (GWRM). In order to predict the (MLP) values, a multi-layer perceptron (MLP) artificial neural network (ANN) has been developed. There are 10 neurons in hidden layer of feed forward (FF) back propagation (BP) network model. The predictive performance of ANN model has been analyzed by comparing the results obtained from the ANN model using Levenberg-Marquard algorithm as the training algorithm with the target values. When the obtained Mean Square Error (MSE), Coefficient of Determination (R) and error rate values have been analyzed, it has been concluded that the ANN model can predict SFC and NN values with high accuracy. According to the findings of current analysis, ANN approach is accurate, effective and conveniently applicable for simulating the slip flow of Williamson fluid towards the stretching plate with heat generation/absorption. The obtained results showed that ANNs are an ideal tool that can be used to predict Skin Friction Coefficients and Nusselt Number values.