Articles

Access the latest knowledge in applied science, electrical engineering, computer science and information technology, education, and health.

Advanced Search
Filter Icon

Filters article

Years

FAQ Arrow
Single Year
Range Year
0
0

Source Title

FAQ Arrow

Authors

FAQ Arrow

31,456 Article Results

Nonlinear backstepping and model predictive control for grid-connected permanent magnet synchronous generator wind turbines

10.11591/ijece.v15i6.pp5091-5105
Adil El Kassoumi , Mohamed Lamhamdi , Ahmed Mouhsen , Mohammed Fdaili , Imad Aboudrar , Azeddine Mouhsen
This research investigates and compares two nonlinear current-control strategies, backstepping control (BSC) and finite control set model predictive control (FCS-MPC) for machine-side and grid-side converters in grid-connected direct-drive permanent magnet synchronous generator (DD-PMSG) wind turbines. Addressing the control challenges in wind energy systems with varying speeds, the study aims to determine which strategy offers superior performance under identical operating conditions. The nonlinear BSC regulates stator and grid currents using Lyapunov-based techniques, while FCS-MPC leverages model predictions to select optimal switching states based on a cost function. A comprehensive simulation using MATLAB/Simulink is conducted, analyzing each controller’s transient behavior, steady-state response, torque ripple, and power quality total harmonic distortion (THD). Results show that FCS-MPC achieves faster convergence, lower overshoot, and superior power quality compared to BSC, though it requires higher computational resources. Statistical validation supports the robustness of FCS-MPC under parameter uncertainties. This work contributes a structured comparison of advanced nonlinear strategies for PMSG-based wind turbines and provides a foundation for future implementations in real-time embedded control systems. Future directions include experimental validation and hybrid model predictive controller- artificial intelligence (MPC-AI) control frameworks.
Volume: 15
Issue: 6
Page: 5091-5105
Publish at: 2025-12-01

Breast cancer detection using ensemble methods

10.11591/ijece.v15i6.pp5633-5646
Alaa Mohamed Ghazy , Hala Bahy Nafea , Fayez Wanis Zaki , Hanan Mohamed Amer
Breast cancer (BC) is one of the most common cancers among women. This study's framework is divided into three phases. Firstly, a majority hard voting approach is used to apply an ensemble classification mechanism as a decision fusion technique on the level of convolutional neural networks (CNNs). Five pre-trained CNNs—visual geometry group 19 (VGG19), densely connected convolutional network 201 (DenseNet201), residual network 50 (ResNet50), mobile network version 2 (MobileNetV2), and inception version 3 (InceptionV3)—are evaluated, using a data splitting test ratio represents 30% of the total dataset. Secondly, the classification results of the five CNNs are compared to get the best-performance model. Then, seven state of art machine classifiers—decision tree (DT), histogram-based gradient boosting classifier (HGB), support vector machine (SVM), random forest (RF), logistic regression (LR), gradient boosting (GB), and extreme gradient boosting (XGB)—are used to improve system performance on the feature vector that was taken from this CNN model. Thirdly, to improve robustness, a majority hard voting technique is used at the external classifier level using the highest four classifiers selected based on their accuracy. Several experiments were conducted in this study, and the results showed that ResNet50 produced the best results in terms of precision and accuracy. The majority voting mechanism improves the system’s accuracy to 99.85% through CNNs and to 100% through traditional classifiers.
Volume: 15
Issue: 6
Page: 5633-5646
Publish at: 2025-12-01

H-shaped terahertz patch antenna with metamaterials for biomedical applications

10.11591/ijece.v15i6.pp5215-5222
Kaoutar Saidi Alaoui , Siraj Younes , Foshi Jaouad
This paper presents the design and simulation of an H-shaped terahertz microstrip patch antenna integrated with a metamaterial (MTM) layer to enhance performance for biomedical sensing applications. The antenna modeled using high frequency structure simulator (HFSS), is optimized for 4.37 THz operation. While FR4 is used in simulations for baseline analysis, alternative low-loss substrates such as polyimide or quartz are recommended for practical THz applications. The antenna design uses an FR4 substrate with a dielectric constant of 4.4 and a thickness of 2 μm. Ground plane, feed line, and patch are made of copper material. The integration of the MTM enhance clearly the antenna characteristics. This integration helps to improve the antenna impedance matching; the reflection coefficients was enhanced from -25.01 to -63.10 dB. Additionally, this integration boost also the antenna radiation characteristics, increasing the gain from 2.62 to 3.86 dB and the directivity from 3.57 to 4.97 dB.
Volume: 15
Issue: 6
Page: 5215-5222
Publish at: 2025-12-01

Simulation and experimental validation of modular multilevel converters capable of producing arbitrary voltage levels using the space vector modulation method

10.11591/ijece.v15i6.pp5234-5248
Tran Hung Cuong , Pham Chi Hieu , Pham Viet Phuong
Modular multilevel converters (MMC) used forDC-AC energy conversion are becoming popular to connect distributed energy systems to the power systems. There are many modulation methods that can be applied to the MMC. The space vector modulation (SVM) method can produce a maximum number of levels, i.e., 2N+1, in which N is the number of sub- modules (SMs) per branch of the MMC. The SVM method can generate rules to apply to MMCs with any number of levels. The goal of this proposal is to easily expand the number of voltage levels of the MMC when necessary while still ensuring the quality requirements of the system. The proposed SVM method only selects the three nearest vectors to generate optimal transition states, therefore making the computations simpler and more efficient. This has reduced the computational load when compared to the previously applied SVM methods. This advantage ensures an optimal switching process and harmonic quality which will significantly improve the effectiveness of the proposed method was demonstrated through simulations on MATLAB/Simulink and experimental tests on 13-levels voltage MMC converter system using a 309 field-programmable gate array (FPGA) kit.
Volume: 15
Issue: 6
Page: 5234-5248
Publish at: 2025-12-01

A hybrid DMO-CNN-LSTM framework for feature selection and diabetes prediction: a deep learning perspective

10.11591/ijece.v15i6.pp5555-5569
Mutasem K. Alsmadi , Ghaith M. Jaradat , Tariq Alsallak , Malek Alzaqebah , Sana Jawarneh , Hayat Alfagham , Jehad Alqurni , Usama A. Badawi , Latifa Abdullah Almusfar
The early and accurate prediction of diabetes mellitus remains a significant challenge in clinical decision-making due to the high dimensionality, noise, and heterogeneity of medical data. This study proposes a novel hybrid classification framework that integrates the dwarf mongoose optimization (DMO) algorithm for feature selection with a convolutional neural network–long short-term memory (CNN-LSTM) deep learning architecture for predictive modeling. The DMO algorithm is employed to intelligently select the most informative subset of features from a large-scale diabetes dataset collected from 130 U.S. hospitals over a 10-year period. These optimized features are then processed by the CNN-LSTM model, which combines spatial pattern recognition and temporal sequence learning to enhance predictive accuracy. Extensive experiments were conducted and compared against traditional machine learning models (logistic regression, random forest, XGBoost), baseline deep learning models (MLP, standalone CNN, standalone LSTM), and state-of-the-art hybrid classifiers. The proposed DMO-CNN-LSTM model achieved the highest classification performance with an accuracy of 96.1%, F1-score of 94.6%, and ROC-AUC of 0.96, significantly outperforming other models. Additional analyses, including confusion matrix, ROC curves, training convergence plots, and statistical evaluations confirm the robustness and generalizability of the approach. These findings suggest that the DMO-CNN-LSTM framework offers a powerful and interpretable tool for intelligent diabetes prediction, with strong potential for integration into real-world clinical decision-support systems.
Volume: 15
Issue: 6
Page: 5555-5569
Publish at: 2025-12-01

Faster R-CNN implementation for hand sign recognition of the Indonesian sign language system (SIBI)

10.11591/ijece.v15i6.pp5759-5769
Paulus Lestyo Adhiatma , Nurcahya Pradana Taufik Prakisya , Rosihan Ariyuana
The Indonesian sign language system (SIBI) is the authorized sign system in Indonesia that the deaf society uses to convey in Indonesian. However, its use still needs to be expanded and more widespread in the community, causing difficulties in communication for hard-of-hearing people. The product of deep learning technologies such as faster region-based convolutional neural network (Faster R-CNN) in object recognition has the potential to help improve communication between deaf people and the general public. This research will implement the Faster R-CNN algorithm with three different residual network (ResNet) architectures (50, 101, and 152) for SIBI recognition. The comparison of the faster R-CNN algorithm with different architectures is also conducted to identify the best architecture for SIBI recognition, and the results are evaluated using accuracy, precision, recall, and F1-score metrics from confusion matrix calculation and execution time. Faster R-CNN model with ResNet-50 architecture showed the best and most efficient performance with accuracy, recall, precision, and F1-score metrics of 96.15%, 95%, 93%, and 94%, respectively, and an execution time of 36.84 seconds in the testing process compared to models with ResNet-101 and ResNet-152 architectures.
Volume: 15
Issue: 6
Page: 5759-5769
Publish at: 2025-12-01

A new algorithm for quality-of-service improvement in mobile ad hoc networks

10.11591/ijece.v15i6.pp5466-5483
Hanafy M. Ali , Adel F. El-Kabbany , Yahia B. Hassan
The quality of service (QoS) in mobile ad hoc networks (MANETs) plays a crucial role in optimizing overall network resource utilization. MANET routing protocols, fundamental to QoS, demand adaptive and swift solutions for efficient path searching. In this context, our paper introduces a novel algorithm based on MANETs, employing a hybrid approach that combines ant colony optimization (ACO) with hybrid multipath quality of service ant (HMQAnt) routing protocols. Our algorithm emphasizes bandwidth optimization as a pivotal factor for providing effective paths. By incorporating bandwidth as a significant parameter in the MANETs algorithm, we aim to enhance its overall properties. The proposed routing protocol, focusing on bandwidth optimization, is anticipated to improve the delivery of total network traffic. Evaluation of the algorithm's performance is conducted through QoS metrics, which are overhead, end-to-end delay, and jitter, throughputs, utilizing a MATLAB simulator. Simulation results indicate that our proposed routing protocol holds a distinct advantage compared to ad hoc on-demand distance vector (AODV), destination- sequenced distance (DSDV), dynamic source routing (DSR), and hybrid ant colony optimization-based (ACO) routing protocol called (ANTMANET) algorithms.
Volume: 15
Issue: 6
Page: 5466-5483
Publish at: 2025-12-01

Plant disease detection and classification: based on machine learning and Eig(Hess)-co-occurrence histograms of oriented gradients

10.11591/ijece.v15i6.pp5336-5346
El Aroussi El Mehdi , Barakat Latifa , Silkan Hassan
Agricultural districts provide high-quality food and contribute substantially to economic growth and population support. However, plant diseases can directly reduce food production and threaten species diversity. The use of precise, automated detection techniques for early disease identification can improve food quality and mitigate economic losses. Over the past decade, numerous methods have been proposed for plant disease classification, and in recent years the focus has shifted toward deep learning approaches because of their outstanding performance. In this study, we employ the Eig(Hess)-co-occurrence histograms of oriented gradients (CoHOG) descriptor alongside pre-trained machine-learning models to accurately identify various plant diseases. We apply principal component analysis (PCA) for dimensionality reduction, thereby enhancing computational efficiency and overall model performance. Our experiments were conducted on the popular PlantVillage database, which contains 54,305 images across 38 disease classes. We evaluate model performance using classification accuracy, sensitivity, specificity, and F1-score, and we perform a comparative analysis against state-of-the-art methods. The findings indicate that the approach we proposed achieves up to 99.83% accuracy, outperforming existing models. Additionally, we test the robustness of our method under various conditions to highlight its potential for real-world agricultural applications.
Volume: 15
Issue: 6
Page: 5336-5346
Publish at: 2025-12-01

Attenuated-chattering global second-order sliding mode load frequency controller for multi-region linked power systems

10.11591/ijpeds.v16.i4.pp2381-2388
Phan-Thanh Nguyen , Cong-Trang Nguyen
In this study, a new chattering-free global second-order sliding mode load frequency controller (CGSOSMLFC) is proposed for multi-region linked power systems (MRLPS). Key achievements of this paper include: i) a new CGSOSMLFC is investigated utilizing only output variables; ii) a global steadiness of the MRLPS is ensured by eliminating the hitting phase in traditional sliding mode control (TSMC), and the undesirable high-frequency vacillation marvel in the control signal is efficiently lessened by utilizing the second-order sliding mode control technique. Firstly, a novel estimator is constructed to conjecture the immeasurable state variables of the MRLPS. Then, an estimator-based CGSOSMLFC is synthesized to force the states of the controlled plant into the anticipated switching surface at an instance time and attenuate the chattering phenomenon in the control indication. Additionally, the total MRLPS’s stability analysis is executed by applying the Lyapunov function theory and linear matrix inequality (LMI), confirming the practicability and reliability of the method. Lastly, simulation outcomes on a three-zone linked power system are furnished to authenticate the usefulness and advantages of the proposed technique.
Volume: 16
Issue: 4
Page: 2381-2388
Publish at: 2025-12-01

Detecting lung nodules in computed tomography images based on deep learning

10.11591/ijece.v15i6.pp5604-5615
Lam Thanh Hien , Le Anh Tu , Pham Trung Hieu , Pham Minh Duc , Nguyen Van Nang , Do Nang Toan
Lung cancer is currently recognized as one of the most dangerous cancers, with high mortality rate. In order to deal with lung cancer, an important task is to detect lung nodules early to improve patient survival rates, and computed tomography (CT) scans are crucial data for this. In this research, we propose a deep learning-based method for detecting lung nodules in the CT images with the goal of increasing the likelihood of nodule appearance in the input data of the network, making it easier for the model to focus on relevant areas while reducing noise from areas unrelated to the result. Specifically, we propose a simple lung region segmentation process and optimize the hyperparameters of the faster region-based convolutional neural networks (faster R-CNN) model based on the analysis of nodule characteristics in CT image data. In our experiments, to evaluate the effectiveness of our proposals, we conducted tests on the standard LUNA16 dataset with different backbone configurations for the model, namely ResNet50, ResNet50v2, and MobileNet. The best results achieved were 0.86 mAP50 and 0.91 Recall for the Resnet50, and 0.84 mAP50 and 0.94 Recall for the ResNet50v2. These impressive outcomes underscore the success of our method and establish a robust basis for future studies to further integrate AI into healthcare solutions.
Volume: 15
Issue: 6
Page: 5604-5615
Publish at: 2025-12-01

Devanagari optical character recognition of printed text

10.11591/ijece.v15i6.pp5914-5923
Malathi P. , Chandrakanth G. Pujari
Hundreds of native languages and scripts are making their way on digital platform to sustain in multiple data formats. Optical character recognition (OCR) is one such dimension where the low resource languages are yet to find their stability. Devanagari OCR is one such low resource script problem to be dealt with, though it is the fourth widely used global script. Recent works carried on OCR have focused on word level approach and face challenges of spiraling complexity as language alphabet set size crosses hundreds. Most of these OCR works are done in constrained environment, with huge datasets and large computational resources. As a result, effective benchmark evaluation of the works against one another on defined metrics is scarce. Aim here is to explore character level Devanagari OCR with printed text images as input. Pattern recognition (PR) principles for diacritic classification and convolutional neural network (CNN) for base character classification are used. word error rate (WER) of 24.47% is attained. However, the training dataset complexity is reduced by 4.35 times. The ten multi class models, training time range from 45 minutes to 2.5 hours. Further the models can be trained in parallel to complete the training process in 3-4 hours. Thus, the approach used for text classification facilitates the Devanagari OCR solution to be offered in off-the-shelf computing devices.
Volume: 15
Issue: 6
Page: 5914-5923
Publish at: 2025-12-01

Fault diagnosis for inverter open circuit faults using DC-link signal and random forest-based technique

10.11591/ijpeds.v16.i4.pp2178-2185
Hoang-Giang Vu , Dang Toan Nguyen
Three-phase voltage source inverters based on insulated-gate bipolar transistors (IGBTs) are widely used in various industrial applications. Faults in IGBTs significantly affect the performance of the inverter and entire system. Robust and accurate fault detection are the key requirements of fault diagnosis methods. This paper explores a method for diagnosing power switch open circuit faults of a voltage source inverter based on machine learning algorithms. The diagnosis is performed in two steps, firstly the fault is detected by applying the Random Forest classifier algorithm with the DC-link signal. Next, the fault switch location is performed by additionally using the inverter output AC current signals. The diagnostic results based on simulation data show that the fault can be detected with maximum accuracy. Meanwhile, the accuracy in locating the fault switch is also significantly improved with the additional use of current signals measured at the DC-link. Potential application of electromagnetic field signal is also highlighted for the practical implementation of fault diagnosis.
Volume: 16
Issue: 4
Page: 2178-2185
Publish at: 2025-12-01

The evolution of routing in VANET: an analysis of solutions based on artificial intelligence and software-defined networks

10.11591/ijece.v15i6.pp5388-5400
Lewys Correa Sánchez , Octavio José Salcedo Parra , Jorge Gómez
This study explored the evolution of vehicular ad hoc networks (VANET) and focused on the challenges and opportunities for routing in these dynamic environments. Despite advancements in traditional protocols, a significant gap persists in the ability to adapt to highly mobile environments with variable traffic, which limits routing efficiency and quality of service. Emerging technologies, such as artificial intelligence (AI) and software- defined networks (SDN), are discussed that have the potential to revolutionize the management of VANET. Machine learning can be used to predict traffic, optimize routes, and adapt routing protocols in real-time. Furthermore, SDN can simplify routing management and enable greater flexibility in network configurations. A comprehensive overview of the convergence of AI and SDN is presented, and the potential complementarities between these technologies to address routing challenges in VANET are explored. Finally, the implications of efficient routing in VANET for road safety, traffic management, and the development of new applications are discussed, and future research lines are identified to address challenges such as scalability, data security, and computational efficiency in vehicular environments.
Volume: 15
Issue: 6
Page: 5388-5400
Publish at: 2025-12-01

Optimized passive and active shielding of magnetic induction generated by ultra-high-voltage overhead power lines

10.11591/ijece.v15i6.pp5144-5161
Salah-Eddine Houicher , Rabah Djekidel , Sid Ahmed Bessidek
This paper presents computational modeling to assess and limit the magnetic induction levels emitted by an extra-high-voltage (EHV) overhead transmission line of 750 kV using the fundamental principle of Biot-Savart law in magnetostatics. An optimization technique based on the grey wolf optimizer (GWO) algorithm is employed to determine the appropriate location of the passive and active loop conductors, and the associated parameters to shielding to achieve better compensation of magnetic induction in an interest zone. The resulting magnetic induction of the ultra high voltage (UHV) overhead power line exhibits a crest value of 27.78 μT at the middle of the right-of-way, which can be considered unacceptable by strict protection standards. Generally, the magnetic compensation loops optimally located under the phase conductors of the power transmission system reduce the magnetic induction levels along the transmission line corridor. The passive loop attenuates the maximum magnetic induction by a rate of 29.7%. Therefore, the performance of the active loop is better; it provides a greater reduction with a rate reaching 53.24%. The simulation results were tested with those derived by the elliptical polarization process. An excellent concordance was found, which made it possible to ensure the adopted method.
Volume: 15
Issue: 6
Page: 5144-5161
Publish at: 2025-12-01

Design and development of home-grown biometric fingerprint device and software for attendance and access control

10.11591/ijece.v15i6.pp5616-5632
Jumoke Soyemi , Ogunyinka Olawale Ige , Olugbenga Babajide Soyemi , Ajibodu Franklin Ademola , Adaramola Ojo Jayeoba , Afolayan Andrew Olumide , Habeeb O. Amode , Mukail Aremu Akinde
This study details the design, development, and deployment of an Android-based Biometric Fingerprint system tailored for institutional access control, attendance tracking, exam monitoring, and staff management. Developed collaboratively by the Innovation Centre and departments across engineering and information and communication technology (ICT), the system integrates custom hardware and software. Hardware includes fingerprint sensors connected to an ATMEGA8 microcontroller and Android interfaces for portability. The software uses modular architecture, comprising a Kotlin-based mobile app with Jetpack Compose, a Laravel-powered web admin panel, and a secure backend API hosted on a virtual private server (VPS). Fingerprint data is safely stored using base64 encoding, enabling accurate user authentication and real-time tracking. A functional prototype was built, tested, and refined, with 95 units deployed in a pilot phase. The system supports multiple fingerprint profiles, secure data handling, and integration with existing institutional platforms. Emphasizing customization, modularity, and adherence to ICT policies, the research also serves as a training tool for staff and students, enhancing operational efficiency and supporting local technology development. Performance evaluation showed a FAR of 0.5%, FRR of 1.2%, and an average authentication time of 2.3 seconds. Post-deployment, student attendance increased by 15%, fee compliance by 10%, and 89% of users rated the system as easy to use. This work demonstrates effective hardware-software co-design for scalable biometric authentication in educational settings.
Volume: 15
Issue: 6
Page: 5616-5632
Publish at: 2025-12-01
1 2 3
...
2097 2098
Show 162 of 2098

Discover Our Library

Embark on a journey through our expansive collection of articles and let curiosity lead your path to innovation.

Explore Now
Library 3D Ilustration