International Journal of Electrical and Computer Engineering Systems

Design and Simulation of Rectangular Slot Antennas Using the Finite Element Method in Python

Zahraa A. Hamza — 2025-09-24

The design and simulation of rectangular slot antennas using a Python-based Finite Element Method (FEM) framework are presented in this study, addressing the limitations of costly and resource-intensive commercial electromagnetic tools and the proposed open-source implementation leverages Python's computational ecosystem—integrating Gmsh for mesh generation, FEniCS for FEM discretization, and SciPy for sparse matrix solving—to provide an accessible and customizable platform for antenna analysis. Validation against Computer Simulation Technology (CST) and High Frequency Structure Simulator (HFSS) demonstrates exceptional agreement, with return loss (S11) deviations below 0.5 dB, radiation efficiencies exceeding 85%, and impedance matching within 2 Ω of the target 50 Ω, parametric studies reveal the impact of slot dimensions and substrate properties on resonant frequency and bandwidth, while computational benchmarks highlight Python-FEM's competitive performance, achieving solve times under 20 seconds for meshes with 180 MB memory usage and the framework's accuracy, coupled with its open-source flexibility, bridges the gap between academic research and industrial prototyping, particularly for applications in 5G, IoT, and radar systems, future enhancements, like Graphics Processing Unit (GPU) acceleration and multi-physical coupling, are proposed to further advance its scalability and versatility in next-generation antenna design.

Analytical Approach to Predict the Magnetic Field of Slotted Permanent Magnet Linear Machines in Open-Circuit mode

Naghi Rostami — 2025-09-22

This paper presents an effective method to calculate the magnetic field distribution of slotted Permanent Magnet Linear Synchronous Machines (PMLSM) with surface-mounted magnets at no-load condition. 2D analytical expressions are employed to make a prediction of magnetic field components. The method proposed in this article has significant advantages in terms of accuracy compared to related studies. More harmonics can be included with new modifications in the analytical calculations. As a result, a more accurate field prediction is obtained. In addition, slot effects are considered in the prediction of the magnetic field. Finite element method (FEM) is used to assess the accuracy of the method.

Optimal Power Control Using Modified Perturb and Observe Algorithm for Photovoltaic System Under Partial Shading

Afrilia Surya Andini — 2025-10-03

Implementation of photovoltaic systems encounters problems, particularly concerning Partial Shading Conditions (PSC), solar irradiance, and temperature, which influence the generated output power. The PSC can diminish the power efficiency of the photovoltaic system. Consequently, a controller is required to optimize the photovoltaic system’s power output by considering the power supply characteristics. This paper discusses optimal power control in photovoltaic system under PSC. The proposed method employs a Modified Perturb and Observe (MP&O) algorithm based on the observation of current and voltage output from the photovoltaic system. The MP&O algorithm is integrated into a microcontroller and will provide PWM signals to operate the synchronous buck converter. Testing was performed under PSC. The experimental results indicated that the synchronous buck converter achieved a performance efficiency of 85%. The efficacy of the MP&O algorithm was evaluated without the MPPT method and conventional P&O algorithm. The MP&O algorithm outperformed compared to without MPPT method and conventional P&O algorithm. The MP&O algorithm yielded more consistent output power and necessitated a quicker tracking duration. The proposed method achieves an average output power efficiency of 84%; in contrast, without the MPPT method, it only reached 57%, and with the conventional P&O algorithm, it attains an efficiency of just 70%.

Distributed Approach to detect DDOS attack based on Elephant Herding Optimization and Pipeline Artificial Neural Network

Yasamin Hamza Alagrash — 2025-10-10

Cybersecurity experts widely acknowledge that a Distributed Denial of Service (DDoS) assault poses a grave threat, capable of inflicting substantial financial losses and tarnishing the reputation of enterprises. Conventional detection methods are insufficient for identifying DDoS attacks. Simultaneously, with their vast potential, machine learning solutions play a vital role in this field. This paper presents a distributed approach for identifying distributed denial-of-service threats using the pipeline artificial neural network method, supported by elephant herding optimization for feature selection and extraction. The proposed artificial neural network pipeline-based model for detecting DDoS attacks comprises several key stages: collecting the dataset, preparing the data, implementing a balanced data strategy, selecting relevant features using the swarm optimization method Elephant Herding Optimization (EHO), training the model, testing its performance, and evaluating its effectiveness. Experimental results demonstrate that the proposed approach effectively enhances DDoS detection accuracy while reducing false positives, making it a promising solution for network security. This model demonstrated a remarkably high ability to detect DDoS attacks with a 99% accuracy. Thorough investigations demonstrate that the model is highly skilled in implementing security measures and reducing the risks connected with emerging security threats. The effectiveness of our proposed solution, leveraging a pipeline method in Artificial Neural Network (ANN), is crucial to building a reliable model, which is evident in its ability to deliver effective results in low complexity. The proposed method achieves 99.99% accuracy, 99.80% precision, and a False Positive Rate (FPR) of 0.002%, outperforming recent models. These results demonstrate the model's superior accuracy and robustness in identifying complex attack patterns while minimizing false positives.

Lettuce Yield Prediction: ElasticNet Regression Model (ElNetRM) for Indoor Aeroponic Vertical Farming System

Gowtham Rajendiran — 2025-09-26

Indoor aeroponic vertical farming systems have revolutionized agriculture by allowing efficient use of space and resources, eliminating the need for soil. These systems improve crop productivity and growth rates. However, accurately predicting lettuce yield in aeroponic environments remains a complex task due to the intricate interactions between environmental, nutrient, and growth parameters. This work aims to address these issues by integrating advanced sensor technologies with ElasticNet Regression Model (ElNetRM) for its hybrid L1 and L2 regularization capabilities, handling multicollinearity and feature selection problems effectively in order to develop a reliable yield prediction framework. The predictive results showcases that the ElNetRM model forecasts lettuce yield with high accuracy of 92% and less error score (RMSE) of 2.28 using a comprehensive dataset from a sensor-equipped indoor aeroponic system. Also, the results demonstrate the superior predictive power of ElNetRM in capturing complex variable relationships, enhancing yield prediction reliability.

Ensemble Deep Learning Approach For Multi- Class Skin Cancer Classification

Ali Abdulameer — 2025-10-07

Skin cancer is one of the most prevalent types of cancer, often caused by prolonged exposure to ultraviolet (UV) radiation, such as sunlight. This cancer is mainly categorized into benign and malignant lesions, where the latter could cause severe complications and even death. Traditional diagnostic methods, such as visual inspection and dermoscopy, often lack accuracy, while biopsy, though highly accurate, is invasive, time-consuming, and costly. This study aims to develop an automated deep learning model that leverages an ensemble of “Convolutional Neural Networks” (CNNs) to perform four-class classification of common skin lesions: Basal Cell Carcinoma (BCC), Benign Keratosis Lesion (BKL), Melanocytic Nevus (NV), and Melanoma (MEL). Seven widely used CNNs in medical imaging, GoogLeNet, InceptionV3, Xception, ResNet18, ResNet50, ResNet101, and DenseNet201, were evaluated for their performance in this classification task. The ISIC2018 and ISIC2019 datasets were employed, and data augmentation techniques were applied to address dataset imbalances. The analysis identified InceptionV3, Xception, and DenseNet201 as the top- performing networks. Therefore, they are utilized for the ensemble model. These networks were used as feature extractors, and their output features were combined and classified using a “Support Vector Machine” (SVM) algorithm. This approach demonstrates the potential of combining CNNs and SVM in an ensemble framework to enhance the accuracy and reliability of automated skin cancer classification. The proposed model achieved an accuracy of 94.46%, outperforming individual CNNs (93.27%) and existing ensemble methods such as Max Voting (94.12%) and hybrid models like DenseNet201 with Random Forest (91.28%).