Multilayer Feedforward Neural Network Research Articles

Automatic Hybrid Access Control in SCADA-Enabled IIoT Networks Using Machine Learning.

The recent advancements in the Internet of Things have made it converge towards critical infrastructure automation, opening a new paradigm referred to as the Industrial Internet of Things (IIoT). In the IIoT, different connected devices can send huge amounts of data to other devices back and forth for a better decision-making process. In such use cases, the role of supervisory control and data acquisition (SCADA) has been studied by many researchers in recent years for robust supervisory control management. Nevertheless, for better sustainability of these applications, reliable data exchange is crucial in this domain. To ensure the privacy and integrity of the data shared between the connected devices, access control can be used as the front-line security mechanism for these systems. However, the role engineering and assignment propagation in access control is still a tedious process as its manually performed by network administrators. In this study, we explored the potential of supervised machine learning to automate role engineering for fine-grained access control in Industrial Internet of Things (IIoT) settings. We propose a mapping framework to employ a fine-tuned multilayer feedforward artificial neural network (ANN) and extreme learning machine (ELM) for role engineering in the SCADA-enabled IIoT environment to ensure privacy and user access rights to resources. For the application of machine learning, a thorough comparison between these two algorithms is also presented in terms of their effectiveness and performance. Extensive experiments demonstrated the significant performance of the proposed scheme, which is promising for future research to automate the role assignment in the IIoT domain.

Kinetic Determination of Acetylsalicylic Acid Using a CdTe/AgInS2 Photoluminescence Probe and Different Chemometric Models

The combination of multiple quantum dots (QDs) in a multi-emitter nanoprobe can be envisaged as a promising sensing scheme, as it enables obtaining a collective response of individual emitters towards a given analyte and allows for achieving specific analyte-response profiles. The processing of these profiles using adequate chemometric methods empowers a more sensitive, reliable and selective determination of the target analyte. In this work, we developed a kinetic fluorometric method consisting of a dual CdTe/AgInS2 quantum dots photoluminescence probe for the determination of acetylsalicylic acid (ASA). The fluorometric response was acquired as second-order time-based excitation/emission matrices that were subsequently processed using chemometric methods seeking to assure the second-order advantage. The data obtained in this work are considered second-order data as they have a three-dimensional size, I × J × K (where I represents the samples' number, J the fluorescence emission wavelength while K represents the time). In order to select the most adequate chemometric method regarding the obtained data structure, different chemometric models were tested, namely unfolded partial least squares (U-PLS), N-way partial least squares (N-PLS), multilayer feed-forward neural networks (MLF-NNs) and radial basis function neural networks (RBF-NNs).

Improving the spaceborne GNSS-R altimetric precision based on the novel multilayer feedforward neural network weighted joint prediction model

Global navigation satellite system-reflection (GNSS-R) sea surface altimetry based on satellite constellation platforms has become a new research direction and inevitable trend, which can meet the altimetric precision at the global scale required for underwater navigation. At present, there are still research gaps for GNSS-R altimetry under this mode, and its altimetric capability cannot be specifically assessed. Therefore, GNSS-R satellite constellations that meet the global altimetry needs to be designed. Meanwhile, the matching precision prediction model needs to be established to quantitatively predict the GNSS-R constellation altimetric capability. Firstly, the GNSS-R constellations altimetric precision under different configuration parameters is calculated, and the mechanism of the influence of orbital altitude, orbital inclination, number of satellites and simulation period on the precision is analyzed, and a new multilayer feedforward neural network weighted joint prediction model is established. Secondly, the fit of the prediction model is verified and the performance capability of the model is tested by calculating the R2 value of the model as 0.9972 and the root mean square error (RMSE) as 0.0022, which indicates that the prediction capability of the model is excellent. Finally, using the novel multilayer feedforward neural network weighted joint prediction model, and considering the research results and realistic costs, it is proposed that when the constellation is set to an orbital altitude of 500 km, orbital inclination of 75° and the number of satellites is 6, the altimetry precision can reach 0.0732 m within one year simulation period, which can meet the requirements of underwater navigation precision, and thus can provide a reference basis for subsequent research on spaceborne GNSS-R sea surface altimetry.

Entry trajectory optimization for hypersonic vehicles based on convex programming and neural network

Trajectory optimization is important in achieving long-range atmospheric entry hypersonic vehicles. However, the trajectory optimization problem for atmospheric entry of hypersonic vehicles is characterized by strong nonlinearity, parameter uncertainties and multiple constraints. This study proposes a novel online trajectory optimization method for hypersonic vehicles based on convex programming and a feedforward neural network. A sequential second-order cone programming (SOCP) method is obtained to describe the trajectory optimization problem after the Gauss pseudo-spectral discretization. Subsequently, multiple optimal trajectories under aerodynamic uncertainties are generated offline and classified as the training and validation datasets. Then, a multilayer feedforward neural network is trained using these datasets and to output the optimal control command online. This method yields approximately 95% shorter computation time compared with the offline SOCP method. Considering the existence of the aerodynamic uncertainties, three terminal states calculated by this method are all smaller than 4.1%. In conclusion, the proposed trajectory optimization method can provide a high-precision, robust entry trajectory for hypersonic vehicles efficiently.

Prediction of hydrogen production by magnetic field effect water electrolysis using artificial neural network predictive models

Developing an efficient water electrolysis (WE) configuration is essential for high-efficiency hydrogen evolution reaction (HER) activity. In this regard, it has been proven that adding a magnetic field (MF) to the electrolysis system greatly improves the hydrogen output rate. In this study, we developed a method based on a machine learning approach to further improve the hydrogen production (HP) system with MF effect WE. An artificial neural network (ANN) model was developed to estimate the effect of input parameters such as MF, electrode material (cathode type), electrolyte type, supplied power (onset voltage), surface area, temperature, and time on HP in different electrolyzer systems. The network was built using 104 experimental data sets from various electrolysis studies. In the study, the percentage contributions of the input parameters to the HP rate and the optimum network architecture to minimize computation time and maximize network accuracy are presented. The model architecture of 7–12–1 was obtained using the best-hidden neurons. The Levenberg-Marquardt (LM) algorithm was used to train the multi-layer feed-forward neural network. Moreover, the utilization of a range of categorical variables to improve ANN prediction accuracy is a significant novelty in this work. Results demonstrated that the output of the trained ANN model fitted well with the experimental data. The test's correlation coefficient (R) and mean squared error (MSE) were 0.973 and 0.01125, respectively, confirming its powerful predictive performance. This ANN application is the first novel viable model to perform prediction using a neural network algorithm in the electrolysis process for MF effect HP using both categorical and continuous data inputs.

Artificial intelligence techniques for encrypt images based on the chaotic system implemented on field-programmable gate array

<span lang="EN-US">Image encryption is an important issue in protecting the content of images and in the area of information security. This article proposes a novel method for image encryption and decryption using the structure of the artificial neural network (ANN)-based chua chaotic system (CCS). This structure was efficiently designed on a field-programmable gate array (FPGA) chip utilizing the xilinx system generator (XSG) tool with the IEEE-754-1985 32-bit floating-point number format. For ANN-based CCS design, a multilayer feed forward neural network (FFNN) structure with three inputs and three outputs was created. This structure consists of one hidden layer with four neurons, each of which has a Tangent Sigmoid activation function. The training of ANN-based CCS yielded a 3.602e-13 mean square error (MSE) value. After successfully training the ANN-based CCS, the design was carried out on FPGA, utilizing the ANN structure's bias and weight values as a reference. The xilinx vivado (2017.4) design suite was used to synthesis and test the ANN-based CCS on the FPGA. The histogram, correlation coefficient, and entropy are used to perform security analysis on various images. Finally, FPGA hardware co-simulation using a Xilinx Artix7 xc7a100t-1csg324 chip was utilized to verify that the encryption and decryption of the images were successful.</span>

Drying Kinetics, Physicochemical and Thermal Analysis of Onion Puree Dried Using a Refractance Window Dryer

Onions have a high moisture content, which makes them more susceptible to microbial growth. Drying is one of the postharvest preservation methods applied to decrease onion moisture content, thereby increasing its storage life. In this study, onions were peeled, washed, cut into quarters, hot water blanched, and pureed. The puree was further dried using two different drying methods: refractance window drying (RWD) (water temperature: 70 °C) and convective drying (CD) (50 °C). The puree was spread on prefabricated trays at varying thicknesses of 2 mm, 4 mm, and 6 mm. It was observed that, irrespective of the drying method, moisture ratio (MR) decreased and drying time and effective moisture diffusivity increased with respect to the thickness of the puree. In addition, the Lewis model and the Wang and Singh model showed the highest R2 and lowest SEE value for RWD and CD, respectively. Moreover, the MR of onion puree during RWD and CD was predicted using a multi-layer feed-forward (MLF) artificial neural network (ANN) with a back-propagation algorithm. The result showed that the ANN model with 12 and 18 neurons in the hidden layer could predict the MR, with a high R2 value for RWD and CD, respectively. The results also showed that the thickness of the puree and drying method significantly affected the physicochemical quality (color characteristics, pyruvic acid content, total phenolic content, total flavonoid content, antioxidant capacity, and hygroscopicity) of onion powder. It was concluded that RWD proved to be a better drying method than CD in terms of the quality of dried powder and reduced drying time. Irrespective of the drying method, 2 mm-thick puree dried yielded the best-dried onion powder in terms of physicochemical quality, as well yielding the lowest drying time. These samples were further analyzed for calculating the glass transition temperature.

INCORPORATING A HONEYFARM WITH MLFFNN IDS FOR IMPROVING INTRUSION DETECTION

Today’s networks must deal with dynamically changing threats each day. Use of static datasets to train and prepare multi-layer feed forward neural network intrusion detection systems (MLFFNN IDS) doesn’t address these new threats. The use of real traffic data to train neural network IDSs has been out of reach in organizations due to privacy and concerns. Now the use of a honeyfarm system can provide real-time data to a MLFFNN IDS so that it can adjust to new threats as they begin. This system also removes the privacy and concerns since information about the network is false and acts as a decoy to lure attackers away from the real organizational network. This paper introduces a honeyfarm architecture one can use with a MLFFNN IDS to improve intrusion detection capability.

Design of motor cable artificial muscle (MC-AM) with tendon sheath–pulley system (TSPS) for musculoskeletal robot

AbstractIn an unstructured environment, the arm can perform complicated tasks with rapidity, flexibility, and robustness. It is difficult to configure multiple artificial muscles similar to an arm in the compact space of a robotic arm. When muscle tension is transferred, mechanisms like tendon-sheath/tendon-pulley may be installed in a compact space to develop musculoskeletal robots that are closer to the arm. However, handling variable frictional nonlinearity and elastic cable deformation is necessary for transmission stability. In this study, the modular artificial muscle system (MAMS), including motor cable artificial muscle and tendon sheath–pulley system (TSPS), that can be installed remotely and transmit muscle tension in narrow paths, is designed. The feed-forward multi-layer neural network (FF-MNN) approach is utilized to discuss the relationship between the measurable input tension of TSPS and the unmeasurable output tension and cable elongation. Subsequently, the lightweight musculoskeletal arm (LM-Arm) is built to verify the validity of MAMS. Through trials, the experiments of MAMS after friction compensating and the LM-Arm’s end-point 3D trajectory tracking are investigated. The results show that average errors of the active and passive muscles tension are 3.87 N and 3.51 N, respectively, under conditions of larger load and higher contraction velocity. The average muscle length error of trajectory tracking is 0.00078 m (0.72%). The suggested MAMS may successfully build a musculoskeletal robot that has similar flexibility and morphology to the arm. It can also be utilized to power various pieces of machinery, such as rescue robot, invasive surgical robots, dexterous hands, and wearable exoskeletons.

MODELING THE TENSILE STRENGTH OF CONCRETE WITH POLYETHYLENE TEREPHTHALATE (PET) WASTE AS REPLACEMENT FOR FINE AGGREGATE USING ARTIFICIAL NEURAL NETWORK

Tensile strength of concrete made with polyethylene terephthalate (PET) waste as replacement for fine aggregate was modelled using artificial neural network. A multilayer feedforward neural network (MLFFNN) and radial basis function (RBF) methodology were compared to see which was more accurate. The MLFFNN modelling results showed a predictive accuracy of 95.364% and a root mean square error value of 4.4409 × 10-16 while RBF neural network modeling results showed a higher predictive accuracy (99.509%) with a lower root mean square error value (1.6653 × 10-16). It is concluded that ANN models accurately predicted the tensile strength of PET concrete.

Deep neural networks for the efficient simulation of macro-scale hysteresis processes with generic excitation waveforms

An effective and performing hysteresis model, based on a deep neural network, with the capability to reproduce the evolution of magnetization processes under arbitrary waveforms of excitation is here presented. The proposed model consists of a standalone multi-layer feed-forward neural network, with reserved input neurons for the past values of both the input (H) and output (M), aiming at the reproduction of the storage mechanism typical of hysteretic systems. The training set has been opportunely prepared starting from a set of simulations, performed by the Preisach hysteresis model. The optimized training procedure, based on multi-stage check of the model performance, will be comprehensively discussed. The comparative analysis between the neural network-based model, implemented at low level of abstraction, and the Preisach model covers additional hysteresis processes, different from those involved in the training. The mild/moderate memory requirement and the significant computational speed make the proposed approach suitable for a future coupling with finite-element analysis.

Machine learning modelling and evaluation of jet fires from natural gas processing, storage, and transport

AbstractJet fires and their repercussions play a significant role in catastrophic incidents that typically have a cascading impact in process industries. Several hydrocarbon experiments from 19 papers were incorporated into the current endeavour to develop simulations of jet flames using machine learning (ML) models. Dimensionless characteristics have been used as output and input variables, including mass flow rates, fuel density, jet flame length, and heat release fluxes. When training three layers of the multi‐layer feedforward neural network (MLFFNN) method, a Bayesian regularization backpropagation approach was adopted and evaluated with the radial based functions (RBF) algorithm. Through an optimization procedure, the first and second hidden layers of the MLFFNN have been optimized to include 10 and five neurons, respectively. The RBF algorithm with 40 neurons in a single layer has been computed using the same method. The best mean square error (MSE) validation results for RBF and MLFFNN were 0.006 and 0.0002, respectively, for 40 and 100 epochs. The MLFFNN and RBF models' respective regression statistical analysis outputs were 0.9949 and 0.9645. The ML method has been identified as a potentially useful technique for precisely predicting the geometrical and radiative characteristics of jet flames.

An Optimal Neural Network for Hourly and Daily Energy Consumption Prediction in Buildings

In this work, hourly and daily energy consumption prediction has been carried out using multi-layer feed forward neural network. The network designed in the proposed architecture has three layers, namely input layer, hidden layer, and output layer. The input layer had eight neurons, output layer had one neuron, and the number of neurons in the hidden layer was varied to find an optimal number for accurate prediction. Different parameters of the neural network were varied repeatedly, and the prediction accuracy was observed for each combination of different parameters to find an optimized combination of different parameters. For hourly energy consumption prediction, a total of six weeks data (September 1 to October 12, 2004) of 10 residential buildings has been used whereas for daily energy consumption prediction, a total of 52 weeks data (January 2004 to December 2004) of 30 residential buildings has been used. To evaluate the performance of the proposed approach, different performance evaluation measurements were applied.

Application of a multi-layer feedforward neural network to predict fracture density in shale oil, Junggar Basin, China

Fengcheng Formation in the Mabei Slope of Junggar Basin has low porosity and permeability. However, fractures are well developed, representing an effective storage space for shale oil. Core and experimental data show that the shale oil reservoir of Fengcheng Formation positively correlates with oil content and fractures. And the fracture density has a good quantitatively positive correlation with crude oil production from the production data. Fengcheng Formation has been significantly enriched and accumulated with shale oil due to fractures serving as reservoirs and seepage channels. Therefore, quantitative prediction of fractures is the key to finding high production areas of shale oil in the Fengcheng Formation. The purpose of this study is to extract the seismic attributes that are sensitive to shale oil reservoir fractures. These attributes include curvature, deep learning fracture detection, maximum likelihood, eigenvalue coherence, and variance cube. Furthermore, a seismic multi-attribute fracture density prediction model is trained at the well point using a feedforward neural network method, and the spatial distribution of fracture density is predicted. The results show that the predicted fracture density is consistent with the formation micro imaging logs in the area. Simultaneously, combined with the understanding of the quantitative relationship between fracture density and shale oil production, quantitative prediction results of fracture density could provide the basis for determining the distribution and optimal location of high-quality shale oil wells in the study area. This study will serve as a benchmark for identifying fractures in shale oil reservoirs worldwide.

A sustainable and secure load management model for green cloud data centres

The massive upsurge in cloud resource demand and inefficient load management stave off the sustainability of Cloud Data Centres (CDCs) resulting in high energy consumption, resource contention, excessive carbon emission, and security threats. In this context, a novel Sustainable and Secure Load Management (SaS-LM) Model is proposed to enhance the security for users with sustainability for CDCs. The model estimates and reserves the required resources viz., compute, network, and storage and dynamically adjust the load subject to maximum security and sustainability. An evolutionary optimization algorithm named Dual-Phase Black Hole Optimization (DPBHO) is proposed for optimizing a multi-layered feed-forward neural network and allowing the model to estimate resource usage and detect probable congestion. Further, DPBHO is extended to a Multi-objective DPBHO algorithm for a secure and sustainable VM allocation and management to minimize the number of active server machines, carbon emission, and resource wastage for greener CDCs. SaS-LM is implemented and evaluated using benchmark real-world Google Cluster VM traces. The proposed model is compared with state-of-the-arts which reveals its efficacy in terms of reduced carbon emission and energy consumption up to 46.9% and 43.9%, respectively with improved resource utilization up to 16.5%.

Predicting the buckling behaviour of thin-walled structural elements using machine learning methods

The design process of thin-walled structural members is highly complex due to the possible occurrence of multiple instabilities. This research therefore aimed to develop machine learning algorithms to predict the buckling behaviour of thin-walled channel elements subjected to axial compression or bending. Feed-forward multi-layer Artificial Neural Networks (ANNs) were trained, in which the input variables comprised the cross-sectional dimensions and thickness, the presence/location of intermediate stiffeners, and the element length. The output data consisted of the elastic critical buckling load or moment, while also providing an immediate modal decomposition of the buckled shape into the traditionally defined ‘pure’ buckling mode categories (i.e. local, distortional and global buckling). The sample output for training was prepared using a combination of the Finite Strip Method (FSM) and the Equivalent Nodal Force Method (ENFM). The ANN models were subjected to a K-fold cross-validation technique and the hyperparameters were tuned using a grid search technique. The results indicated that the trained algorithms were capable of predicting the elastic critical buckling loads and carrying out the modal decomposition of the critical buckled shapes with an average accuracy (R2-value) of 98%. The influence of the various channel parameters on the output was assessed using the SHapley Additive exPlanations (SHAP) method.

Responses, optimization and prediction of energy harvesters under galloping and base excitations

In this paper, the complexification-averaging method is firstly used to derive the theoretical solutions and analyze the dynamical performance of the energy harvesters under galloping and base excitations. Based on the distributed parameter model of the energy harvesters, the model equations can be turned into a set of high-order ordinary differential equations by solving the ordinary differential equations with coupled terms, then the equation of the amplitude frequency and the analytical relationship to determine the displacement amplitude and the voltage amplitude are derived. In addition, the influence of parameters such as the wind speed, the electromechanical coupling parameter and the base excitation frequency on the nonlinear behaviors of the energy harvesters are explored. It is found that there exists very good consistency between the theoretical results and the numerical results. The response properties are verified by means of the phase portraits and time histories. Additionally, according to the implicit function derivative theorem, the critical condition for identifying the unstable region is obtained. And with the different system parameters, the critical unstable boundary can be identified. In order to improve the efficiency of energy harvesting, the parameters of the energy harvesters are optimized based on the purpose of maximum voltage amplitude. Finally, the results are predicted by means of the multilayer feedforward neural network, and the excellent prediction results can be found.

Development of surrogate models for evaluating energy transfer quality of high-speed railway pantograph-catenary system using physics-based model and machine learning

High-speed railway pantograph-catenary system is the only energy transfer pathway to drive a train operation. Energy transfer quality deteriorates with the increasing train speed and harsh service environment, thereby quickly and accurately evaluating the energy transfer quality is very important to guarantee the normal operation of a train. In this study, firstly, the physics-based model to simulate the dynamic interaction of pantograph-catenary system is established and validated. Eleven input parameters involve the essential line design and train operation parameters, and the output parameters that are crucially responsible for energy transfer quality are obtained by feature extraction. Secondly, a sampling strategy is employed to construct the input sampling points, based on which the outputs are computed via physics-based model, then combining them the dataset is obtained. Thirdly, five tree-based classification surrogate models are developed and compared to assess the level of energy transfer quality. Finally, eight regression surrogate models are developed in replacing physics-based model to evaluate the essential values of energy transfer quality. It is found that the gradient boosting decision tree (GBDT)-based surrogate model is the optimal classification model and the multi-layer feed-forward deep neural network (MLF-DNN)-based surrogate model for the optimal regression model. The two surrogate models are expected to quickly find the optimal design parameters and improve the operation control of trains of high-speed railway for the purpose of enhancing the energy transfer quality if coupled with optimization procedure.

Comparative efficacy of three different methods for prediction of live body weight in small ruminants

The present investigation was carried out on 216 records of Osmanabadi goat and 208 records of Deccani sheep. The relative efficiency of three different methods, viz. Shaffer's formula method, multiple linear regression and artificial neural network for prediction of live body weight were investigated. Individual animals were weighed on electronic weighing balance along with their body measurements like lengths, height and girths were measured. The explanatory variables were body lengths, body height and chest girths while dependant variable was body weight. A multilayer feed forward neural network with back propagation of error learning mechanism was developed using artificial neural network using bayesian regularization algorithms. It was observed that artificial neural network was best fitted with in goat and sheep, with the adjusted R2 of 0.93, explained by its linear relationship with the explanatory variables in goat. However, the prediction accuracy (R2 value) was observed as 94.21% with 2.35 kg error. While in sheep, the adjusted R2 was 0.82 and the prediction accuracy (R2 value) was observed as 85.29% with 3.48 kg error. The multiple linear regressions observed the adjusted R2 of 0.894, and the prediction accuracy (R2 value) as 90.03% with 4.73 kg error. The correlation coefficients for different body measurements using three different methods were ranged from 0.952 to almost one.

Prediction of body condition in Jersey dairy cattle from 3D-images using machine learning techniques.

The body condition of dairy cows is a crucial health and welfare indicator that is widely acknowledged. Dairy herds with a well-management body condition tend to have more fertile and functional cows. Therefore, routine recording of high-quality body condition phenotypes is required. Automated prediction of body condition from 3D images can be a cost-effective approach to current manual recording by technicians. Using 3D-images, we aimed to build a reliable prediction model of body condition for Jersey cows. The dataset consisted of 808 individual Jersey cows with 2,253 phenotypes from three herds in Denmark. Body condition was scored on a 1 to 9 scale and transformed into a 1 to 5 scale with 0.5-unit differences. The cows' back images were recorded using a 3D camera (Microsoft Xbox One Kinect v2). We used contour and back height features from 3D-images as predictors, together with class predictors (evaluator, herd, evaluation round, parity, lactation week). The performance of machine learning algorithms was assessed using H2O AutoML algorithm (h2o.ai). Based on outputs from AutoML, DeepLearning (DL; multi-layer feedforward artificial neural network) and Gradient Boosting Machine (GBM) algorithms were implemented for classification and regression tasks and compared on prediction accuracy. In addition, we compared the Partial Least Square (PLS) method for regression. The training and validation data were divided either through a random 7:3 split for 10 replicates or by allocating two herds for training and one herd for validation. The accuracy of classification models showed the DL algorithm performed better than the GBM algorithm. The DL model achieved a mean accuracy of 48.1% on the exact phenotype and 93.5% accuracy with a 0.5-unit deviation. The performances of PLS and DL regression methods were comparable, with mean coefficient of determination of 0.67 and 0.66, respectively. When we used data from two herds for training and the third herd as validation, we observed a slightly decreased prediction accuracy compared to the 7:3 split of the dataset. The accuracies for DL and PLS in the herd validation scenario were > 38% on the exact phenotype and > 87% accuracy with 0.5-unit deviation. This study demonstrates the feasibility of a reliable body condition prediction model in Jersey cows using 3D-images. The approach developed can be used for reliable and frequent prediction of cows' body condition to improve dairy farm management and genetic evaluations.