Artificial Neural Network-based Method Research Articles

A new understanding of the alkali-silica reaction expansion in concrete using a hybrid ensemble model

Alkali-Silica Reaction (ASR) is a critical issue affecting the durability of concrete structures, leading to significant long-term maintenance costs and safety concerns. This study addresses the need for accurate prediction of ASR-induced expansion in concrete, leveraging advanced machine learning techniques. Two hybrid ensemble methods, the inverse variance method and the artificial neural network-based ensemble method, were proposed to integrate the individual models to enhance the predictive capabilities. A comprehensive experimental database comprising over 1997 sets of ASR expansion data was collected from relevant literature with the environmental conditions, specimen geometry and material composition as the input variables and specimen strain as the output variable. The findings indicate that the predictive performance of individual models falls short of the performance of the hybrid ensemble machine learning model, with the inverse variance-based ensemble model exhibiting the highest performance (R = 97.2 %, RMSE = 0.066 mm). The Shapley Additive Explanations analysis reveals that the most influential factors include reaction days, alkali content, aggregate particle size, and silica content, contributing to up to 35 % of the variation in ASR expansion. This work offers valuable insights for the efficient assessment of durability of concrete structures concerning ASR expansion.

Artificial neural network-based method for overhead lines magnetic flux density estimation

Artificial neural network-based method for overhead lines magnetic flux density estimation

Subject-specific trunk segmental masses prediction for musculoskeletal models using artificial neural networks.

Accurate determination of body segment parameters is crucial for studying human movement and joint forces using musculoskeletal (MSK) models. However, existing methods for predicting segment mass have limited generalizability and sensitivity to body shapes. With recent advancements in machine learning, this study proposed a novel artificial neural network-based method for computing subject-specific trunk segment mass and center of mass (CoM) using only anthropometric measurements. We first developed, trained, and validated two artificial neural networks that used anthropometric measurements as input to predict body shape (ANN1) and tissue mass (ANN2). Then, we calculated trunk segmental mass for two volunteers using the predicted body shape and tissue mass. The body shape model (ANN1) was tested on 279 subjects, and maximum deviation between the predicted body shape and the original was 28mm. The tissue mass model (ANN2) was evaluated on 223 subjects, which when compared to ground truth data, had a mean error of less than 0.51% in the head, trunk, legs, and arms. We also compared the two volunteer's trunk segment mass with experimental data and found similar trend and magnitude. Our findings suggested that the proposed method could serve as an effective and convenient tool for predicting trunk mass.

Genetic Discrimination of Grade 3 and Grade 4 Gliomas by Artificial Neural Network.

Gliomas, including anaplastic gliomas (AG; grade 3) and glioblastomas (GBM; grade 4), are malignant brain tumors associated with poor prognosis and low survival rates. Current classification systems based on histopathology have limitations due to intratumoral heterogeneity. The treatment and prognosis are distinctly different between grade 3 and grade 4 gliomas patients. Therefore, there is a need for molecular markers to differentiate these tumors accurately. In this study, we aimed to identify a gene expression signature using an artificial neural network (ANN) in application to microarray and serial analysis of gene expression (SAGE) data for grade 3 (AG) and grade 4 (GBM) gliomas discrimination. We acquired gene expression data from publicly available datasets on glial tumors of grades 3 and 4-a total of 93 grade 3 gliomas and 224 grade 4 gliomas. To select genes for classification, we implemented an artificial neural network-based method using a combination of self-organized maps (SOM) and perceptron. In general, we implemented a multi-stage procedure that involved multiple runs of a genetic algorithm to identify genes that provided optimal clusterization on the SOM. We performed this procedure multiple times, resulting in different sets of genes each time. Eventually, we selected several genes that appeared most frequently in the reduced sets and performed classification using them. Our analysis identified a set of seven genes (BCAS4, GLUD2, KCNJ10, KCND2, AKR7A2, FOLR1, and KIAA0319). The classification accuracy using this gene set was 87.5%. These findings suggest the potential of this gene set as a molecular marker for distinguishing grade 3 (AG) from grade 4 (GBM) gliomas.

Prediction for blood lactate during exercise using an artificial intelligence-Enabled electrocardiogram: a feasibility study.

Introduction: The acquisition of blood lactate concentration (BLC) during exercise is beneficial for endurance training, yet a convenient method to measure it remains unavailable. BLC and electrocardiogram (ECG) both exhibit variations with changes in exercise intensity and duration. In this study, we hypothesized that BLC during exercise can be predicted using ECG data. Methods: Thirty-one healthy participants underwent four cardiopulmonary exercise tests, including one incremental test and three constant work rate (CWR) tests at low, moderate, and high intensity. Venous blood samples were obtained immediately after each CWR test to measure BLC. A mathematical model was constructed using 31 trios of CWR tests, which utilized a residual network combined with long short-term memory to analyze every beat of lead II ECG waveform as 2D images. An artificial neural network was used to analyze variables such as the RR interval, age, sex, and body mass index. Results: The standard deviation of the fitting error was 0.12mmol/L for low and moderate intensities, and 0.19mmol/L for high intensity. Weighting analysis demonstrated that ECG data, including every beat of ECG waveform and RR interval, contribute predominantly. Conclusion: By employing 2D convolution and artificial neural network-based methods, BLC during exercise can be accurately estimated non-invasively using ECG data, which has potential applications in exercise training.

A Literature Review on the Optimal Placement of Static Synchronous Compensator (STATCOM) in Distribution Networks

The existing distribution networks were designed at a time when there was virtually no embedded generation. The design methods ensured the voltage at various parts of the network remained within the limits required by standards, and for the most part, this was very successfully achieved. As Distributed Energy Resources (DERs) started to grow, the rise in voltage due to injected currents and the local impedances started to push network voltages toward, and even above, the desired upper limits. Voltage limits are based on typical appliance requirements, and long-term over-voltages will ultimately result in unacceptably short appliance life spans. Distribution Static Compensators (dSTATCOMs) are shunt-connected devices that can improve low-voltage networks’ performance by injecting currents that do not transfer real power. The currents can be reactive, negative or zero sequence, or harmonic. System performance can be improved by reducing conduction loss, improving voltage profile and voltage balance, or reducing Total Harmonic Distortion (THD). To obtain these benefits, optimal sizes of dSTATCOMs need to be placed at optimal locations within the distribution network. This paper has considered seventy research articles published over the past years related to the optimal placement and sizing of dSTATCOMs. In this study, minimization of power losses, voltage profile improvement, loadablity factor, voltage sag mitigation, and reduction in annual operating costs are considered fitness functions that are subjected to multiple constraint sets. The optimization algorithms found in the literature are categorized into six methods: analytical methods, artificial neural network-based methods, sensitivity approaches, metaheuristic methods, a combination of metaheuristic and sensitivity analysis, and miscellaneous. This study also presents a comparison among distribution network types, load flow methods optimization tools, etc. Therefore, a comprehensive review of optimal allocation and sizing of dSTATCOMs in distribution networks is presented in this paper, and guidance for future research is also provided.

Artificial neural network based load balancing scheme for top of rack switches in optical data centers

Abstract Data centers serve as dedicated facilities for housing computer systems and their related components, including telecommunications and storage systems. They typically have high levels of security and environmental controls to ensure that the equipment housed within them functions optimally. Data center networks (DCNs) often employ load balancing algorithms to handle large volumes of traffic and ensure that all servers and switches are utilized equally, keeping the network running smoothly. However, as load on the server varies, therefore dynamic traffic management systems that can adjust traffic flow in real-time based on the current traffic state is required. This study presents an artificial neural network-based load balancing method. By training a feed-forward artificial neural network (ANN) using a back propagation (BP) learning algorithm, it evenly distributes workload over all of the nodes. Simulation results are also presented to prove the usefulness of the proposed load balancing mechanism. It is found that the load balancing scheme can reduce the packet blocking probability (PBP) by 10 folds and delay by about nearly 11 percent.

Predicting carbonation depth of concrete using a hybrid ensemble model

This study aims to develop a robust and accurate machine learning-based model to improve the prediction accuracy of carbonation depth in complex concrete structures. Two hybrid ensemble methods, the inverse variance method and the artificial neural network-based ensemble method, were proposed to integrate multiple algorithms. The models were trained on a dataset of 532 data points with 6 input variables. Performance evaluation metrics and the Taylor diagram were used to compare the models. The results showed that the hybrid ensemble models outperformed the single models, with the inverse variance-based model achieving the highest performance (R = 0.975, RMSE = 2.978 mm). Additionally, the contribution analysis indicated that carbonation time, CO2 concentration, and the amount of binder were the most influential factors. This study provides an efficient prediction model for carbonation depth and valuable insights for carbonation-resistant design in concrete structures.

Correction of global digital elevation models in forested areas using an artificial neural network-based method with the consideration of spatial autocorrelation

ABSTRACT To remove vegetation bias (VB) from the global DEMs (GDEMs), an artificial neural network (ANN)-based method with the consideration of elevation spatial autocorrelation is developed in this paper. Three study sites with different forest types (evergreen, mixed evergreen-deciduous, and deciduous) are employed to evaluate the performance of the proposed model on three popular 30-m GDEMs, including SRTM1, AW3D30, and COPDEM30. Taking LiDAR DTM as the ground truth, the accuracy of the GDEMs before and after VB correction is assessed, as well as two existing GDEMs including MERIT and FABDEM. Results show that all the original GDEMs significantly overestimate the LiDAR DTM in the three forest types, with the largest biases of 21.5 m for SRTM1, 26.3 m for AW3D30, and 27.18 m for COPDEM30. Taking data randomly sampled from the corrected area as the training points, the proposed model reduces the mean errors (root mean square errors) of the three GDEMs by 98.8%−99.9% (55.1%−75.8%) in the three forests. When training data have the same forest type as the corrected GDEM but under different local situations, the proposed model lowers the GDEM errors by at least 76.9% (44.1%). Furthermore, our corrected GDEMs consistently outperform the existing GDEMs for the two cases.

Consumption Forecasting and Economic-Financial Evaluation of a Brazilian Company in the Free Market

The essential difference between the Free Contracting Environment (FCE) and the Regulated Contracting Environment (RCE) is the possibility of freely negotiating energy terms and prices with suppliers. Disconnected from the tariffs regulated by the government, in the FCE, consumers bear the costly difference between the contracted energy and that consumed. This cost can be reduced with accurate knowledge of the consumer profile, based on the analysis of historical data. In this article, a methodology is proposed to evaluate the migration of consumers to the FCE. In a case study, graphical statistical techniques help identify the profile of a consumer in the city of Rio de Janeiro, subgroup A4 and with green tariff modality, in the period from 2016 to 2019. Then, classical and artificial neural network-based methods are used for consumption forecasting twelve months ahead. In particular, Long and Short Term Memories (LSTM) networks performed better than Autoregressive Integrated Moving Average (ARIMA) models. At the end, it is demonstrated with economic and financial indicators, the right decision of this consumer to migrate to the FCE, prior to the analysis performed in this case study.

Automatic Extraction of Flooding Control Knowledge from Rich Literature Texts Using Deep Learning

Flood control is a global problem; increasing number of flooding disasters occur annually induced by global climate change and extreme weather events. Flood studies are important knowledge sources for flood risk reduction and have been recorded in the academic literature. The main objective of this paper was to acquire flood control knowledge from long-tail data of the literature by using deep learning techniques. Screening was conducted to obtain 4742 flood-related academic documents from past two decades. Machine learning was conducted to parse the documents, and 347 sample data points from different years were collected for sentence segmentation (approximately 61,000 sentences) and manual annotation. Traditional machine learning (NB, LR, SVM, and RF) and artificial neural network-based deep learning algorithms (Bert, Bert-CNN, Bert-RNN, and ERNIE) were implemented for model training, and complete sentence-level knowledge extraction was conducted in batches. The results revealed that artificial neural network-based deep learning methods exhibit better performance than traditional machine learning methods in terms of accuracy, but their training time is much longer. Based on comprehensive feature extraction capability and computational efficiency, the performances of deep learning methods were ranked as: ERNIE > Bert-CNN > Bert > Bert-RNN. When using Bert as the benchmark model, several deformation models showed applicable characteristics. Bert, Bert-CNN, and Bert-RNN were good at acquiring global features, local features, and processing variable-length inputs, respectively. ERNIE showed improved masking mechanism and corpus and therefore exhibited better performance. Finally, 124,196 usage method and 8935 quotation method sentences were obtained in batches. The proportions of method sentence in the literature showed increasing trends over the last 20 years. Thus, as literature with more method sentences accumulates, this study lays a foundation for knowledge extraction in the future.

Nonparametric estimation-based five-layer neural network RAIM with improved availability

The monitoring performance of receiver autonomous integrity monitoring (RAIM) is restricted when visible satellites are limited in challenging environments. For that, artificial neural network-based RAIM methods have been investigated to improve the detection efficacy. Nevertheless, their corresponding fault exclusion and protection level algorithms are hardly provided for integrity assessments. In this regard, a nonparametric estimation-based RAIM method (NE-RAIM) is investigated to support fault detection, exclusion, and protection level calculation in this paper, boosting the declined monitoring capacity caused by the decrease of visible satellites. We propose a classification variable and a dynamic sampling method based on the variance inflation theory and then obtain the regression of the classification variable using nonparametric estimation. In this way, a five-layer NE-RAIM neural network is constructed to enhance the detection capability further. We also provide a NE-RAIM-based fault exclusion strategy by analyzing the detection result vector. Meanwhile, a protection level algorithm is proposed to enable direct integrity and availability evaluation based on searching the worst-case scenario where the missed detection risk is maximized. Results show that NE-RAIM requires a minimum pseudorange bias of 35 m to realize 100% detection rates under all single-faulty-satellite modes. Compared with least-square RAIM and advanced RAIM, NE-RAIM improves overall 24 h availability by 59.30% and 4.52%, respectively.

Artificial Neural Network-Based Pole-Tracking Method for Online Stabilization Control of Grid-Tied VSC

To cope with the weak grid stability issue of grid-tied voltage source converters (VSCs), this article proposes an artificial neural network (ANN) based approach for online stabilization control of the grid-tied VSC with the pole-tracking feature. First, an ANN is adopted to establish the mapping between the control parameters and the closed-loop poles of the grid-VSC system, serving as a computationally light model surrogate that is favorable for real-time control applications. Then, an online parameter search algorithm enabling simultaneous tuning of multiple controllers and parameters is developed, by which the system's poles under a new grid condition can be pulled to the reference ones, i.e., achieving the pole-tracking-based stabilization control of this article. Finally, the efficacy of the proposed method along with its stabilization effect is verified by MATLAB simulations and experimental results.

Nature-inspired self-organizing collision avoidance for drone swarm based on reward-modulated spiking neural network

Biological systems can exhibit intelligent swarm behavior through relatively independent individual, local interaction and decentralized decision-making. A major research challenge of self-organized swarm intelligence is the coupling influences between individual behaviors. Existing methods optimize the behavior of multiple individuals simultaneously from a global perspective. However, these methods lack in-depth inspiration from swarm behaviors in nature, so they are short of flexibly adapting to real multi-robot online decision-making tasks. To overcome such limits, this paper proposes a self-organized collision avoidance model for real drones incorporating a bio-inspired reward-modulated spiking neural network (RSNN). The local interaction and autonomous learning of a single individual leads to the emergence of swarm intelligence. We validated the proposed model on swarm collision avoidance tasks (a swarm of unmanned aerial vehicles without central control) in a bounded space, carrying out simulation and real-world experiments. Compared with artificial neural network-based online learning methods, our proposed method exhibits superior performance and better stability.

Artificial Neural Network-based Finite Element method for assessing fatigue and stability of an origami-inspired structure

In this paper, we present a comprehensive study of mechanical characteristics of a reconfigurable origami-inspired structure using Finite Element and Artificial Neural Network (ANN) approaches. We introduce a design of crease section in the proposed reconfigurable structure, which undergoes enormous and complicated deformation in the folding and unfolding process. Although this crease design can make the deploying process more straightforward, it can jeopardize the stability and life cycle of the structure. We explore how the geometric parameters of the design affect the stability and fatigue failure, including length ratio, total height, story’s height, thickness, crease indexes, and circumscribed circle’s radius. In order to reduce the computational time, we develop an ANN employing the obtained Finite Element method (FEM) results. The ANN results demonstrate that decreasing the circumscribed circle’s radius, the length ratio, and the total height enhance the stability of the origami-inspired structure. It is found that crease indexes affect stability based on the radius of the circumscribed circle. In addition, we investigate how these parameters simultaneously contribute to this design’s buckling load and life cycle. The results provide a detailed design parameter characteristic map that can be used to optimize origami structure performance.

Application of Deep Learning Technology in the Recommendation System of Constitutional and Constitutional Cases

Due to the huge system in the field of constitution, the content of cases is complicated, and the number of documents related to constitutional cases is huge. It is incredibly tough to find the information that suits your demands in such a large volume of constitutional documents. The artificial neural network-based deep learning method has made significant progress. Deep learning has been used in the disciplines of computer image and video processing, as well as natural language processing, and has outperformed typical machine learning methods. To that purpose, the content-based recommendation algorithm in this work is improved using deep learning approaches. This paper primarily covers the following topics: (1) the current research status of deep learning technology in recommender systems is reviewed, and the theoretical knowledge of deep neural networks is introduced. (2) A text classification model OCCNN based on character-level convolutional neural network is proposed to solve the problem of document classification of massive cases. A semantic similarity calculation model TF-W2V based on word frequency and word vector is also proposed to solve the problem of semantic similarity calculation of cases. (3) The two methods are tested, and the experimental results show that the accuracy is much higher than that of the conventional method. The model can recommend cases that meet the needs of users according to the requirements put forward by users in the process of handling cases.

High resolution in non-destructive testing: A review

Since the beginning of the applications of non-destructive testing/evaluation (NDT/NDE) techniques, efforts have been made consistently to improve their detection sensitivity and resolution. In the present paper, factors governing lateral resolution in three major NDT techniques, viz., ultrasonic testing (UT), x-ray radiographic testing (XRT), and eddy current testing (ECT) are presented. Furthermore, a review of recent advances in these NDT techniques to reach the theoretically achievable resolution limit or even surpassing the same using alternate approaches is also discussed. For example, resolution in UT is theoretically limited to half the wavelength by the Rayleigh limit; however, subwavelength resolutions have been achieved through the applications of near field methods by capturing the evanescent field. On the other hand, the resolution achieved in XRT is primarily limited to half the source/focal spot size, which is many orders of magnitude larger than the wavelength. Over the years, the reduction in the focal spot from macro-focus to micro-focus and now to nano-focus has led to improvement in the resolution to a few nanometers, of course, in combination with suitable magnification required due to detectors with limited pixel size (a few μm to a few 10 s of μm). Similarly, innovations in electromagnetic/magnetic sensors have significantly improved the resolution achieved in ECT. Atomic force microscopy, metamaterials, and artificial neural network-based methods have been employed for obtaining high-resolution NDE images. At the end, authors' perspective toward possible directions for high-resolution NDT is presented.

Deep Learning-Based Subtask Segmentation of Timed Up-and-Go Test Using RGB-D Cameras.

The timed up-and-go (TUG) test is an efficient way to evaluate an individual’s basic functional mobility, such as standing up, walking, turning around, and sitting back. The total completion time of the TUG test is a metric indicating an individual’s overall mobility. Moreover, the fine-grained consumption time of the individual subtasks in the TUG test may provide important clinical information, such as elapsed time and speed of each TUG subtask, which may not only assist professionals in clinical interventions but also distinguish the functional recovery of patients. To perform more accurate, efficient, robust, and objective tests, this paper proposes a novel deep learning-based subtask segmentation of the TUG test using a dilated temporal convolutional network with a single RGB-D camera. Evaluation with three different subject groups (healthy young, healthy adult, stroke patients) showed that the proposed method demonstrated better generality and achieved a significantly higher and more robust performance (healthy young = 95.458%, healthy adult = 94.525%, stroke = 93.578%) than the existing rule-based and artificial neural network-based subtask segmentation methods. Additionally, the results indicated that the input from the pelvis alone achieved the best accuracy among many other single inputs or combinations of inputs, which allows a real-time inference (approximately 15 Hz) in edge devices, such as smartphones.

Adaptive Multi-Parameter-Tuning for Online Stabilization Control of Grid-Tied VSC: An Artificial Neural Network-Based Method

Voltage Source Converter (VSC) for grid integration of renewable energies are prone to have small-signal stability issues when connected to weak AC grids. Such stability issues largely arise from the lack of VSC control adaptivity to the varying grid condition (e.g., grid impedance). To address this issue, this letter presents an adaptive multi-parameter tuning method using the Artificial Neural Network. Innovative aspect of the proposal lies in that it enables the VSC to simultaneously tune multiple controller parameters online, which brings about a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">pole-tracking</i> -based stabilization control feature for the VSC. Experimental results demonstrate that the proposed method can effectively and adaptively stabilize the VSC when the grid impedance is varied.

Development of an artificial neural network-based method for determining the flexibility of power systems with high share of wind generation

A method for the online determination of the resilience of an electric power system using artificial neural networks having various structures is presented. A developed algorithm comprised of an artificial neural network with multiple learning paradigms is used for the rapid calculation of the adaptability index of the electric power system. A satisfactory time for obtaining results is ensured by dividing the adaptability calculation into offline and online processes. To train the neural networks, various methods were used. The multilayer perceptron was trained using the method of back-ward propagation of error, while training of the Kohonen neural network was performed based on the winner-take-all rule. Euclidean distance was used as a proximity measure between the studied vectors. An algorithm for analysing the results obtained by two types of artificial neural networks having dissimilar structures was developed in order to select their optimal structure and recommend a neural network for the real-time determination of the resilience of an electric power system. The proposed algorithm was validated on a 6-node scheme following the command script: computing the resilience of a given system, functioning in multiple modes. The criterion analysis showed that the structures of multilayer perceptron having 16 neurons in a hidden layer and Kohonen neural network having 9 output neurons represent the optimal solution for determining the steady-state mode at the minimum resilience in real-time. According to the results, the value of the resilience of the system varies over the course of a day. The possibility of using artificial neural networks for determining the resilience of electric power systems in real-time is demonstrated.