Vol. 4 No. 2 (2025): August 31
Full Issue
Research Articles
-
Article: 10Published: 31.08.2025 pp. 1-14
Estimating Solar Power Generation with RF, GB, and SVR Algorithms Based on Meteorological Data and Orientation Angles: Adıyaman Case Study
- 505Views 390PDF access
AbstractThis study attempts to develop precise machine learning algorithms for estimating solar power generation using meteorological data, including air pollution (PM2.5, PM10) for different orientation angles in the dual-axis solar panel tracking system. Our study focused on maximizing energy generation for multivariable input factors in Adıyaman, rather than optimizing environmental parameters for fixed panels as previously studied. Three machine learning algorithms—Random Forest (RF), Gradient Boosting (GB), and Support Vector Regression (SVR)—were studied. The experimental setup was designed to measure power, temperature, humidity, wind speed/direction, pressure, air pollution data, and to adjust orientation angles (tilt and azimuth angles). Algorithm performances were assessed using R², MSE, and RMSE metrics. RF yielded the best results (R²: 0.83, MSE: 5.32, RMSE: 2.26). Including cumulative air pollution data improved prediction accuracy, implying particulate matter indirectly affects solar radiation and energy output.
-
Article: 14Published: 31.08.2025 pp. 15-27
A Comparative Study on Wavelet and EMD Methods for Remote detection of Rotational Imbalance
- 490Views 466PDF access
AbstractThis paper presents a radar-driven method for detecting rotational imbalance in machinery using multi-resolution signal processing and machine learning. Two feature extraction strategies—Empirical Mode Decomposition (EMD) and Wavelet Decomposition (WD)—are compared with WD evaluated using Haar, Coiflet (coif2), and Symlet (sym4) mother wavelets. Four classifiers—Support Vector Machine (SVM), k-Nearest Neighbors (KNN), Random Forest (RF), and Decision Tree (DT)—are used to assess performance. Radar signals undergo multi-resolution analysis to derive features that capture imbalance signatures. The performances of the methods are evaluated through classification accuracy, computational cost, and real-time feasibility. Wavelet-based features consistently outperform EMD across all classifiers and wavelets; Haar delivers the best accuracy of 95.28% with RF. A notable overall trade-off is achieved by Haar features with SVM, providing high accuracy with low latency (~219 ms training, ~13.5 ms prediction), while DT provides the fastest inference (0.6 ms) for strict real-time needs. These results provide a practical framework and clear guidelines for algorithm selection in industrial radar-based condition monitoring.
-
Article: 12Published: 31.08.2025 pp. 28-35
Improving Performance of Squirrel Cage Induction Motor Used in Submersible Pumps
- 553Views 497PDF access
AbstractThis study examines the design of a three-phase induction motor used in submersible pumps. The motor considered is a 380V, 50Hz, 75Hp, 2-pole, 8" three-phase induction motor. The targeted induction motor performance is high efficiency at the rated load. Furthermore, the operating characteristics of the three-phase induction motor, such as starting current, starting torque, breakdown current, breakdown torque, rated current, rated torque, and magnetic flux distribution, were also investigated. The design process began with determining the general geometry of the motor and continued with the calculation of the stator, rotor, and equivalent circuit parameters. Geometric and magnetic constraints were considered during the design process, and priority was given to ensuring that the motor achieves the desired high performance. The motor design simulation was first performed in the Ansys RMxprt module, and then the motor design was simulated in the Ansys Maxwell 2D module and compared with motors from various manufacturers. The resulting motor design geometry has a 4% higher efficiency. Other performance outputs of the motor are also found to be sufficient.
-
Article: 11Published: 31.08.2025 pp. 36-42
Capacitive Sensor System for Measuring Moisture Content in Transformer Oil
- 539Views 697PDF access
AbstractThe efficiency and long-term reliability of electrical transformers are highly dependent on the quality of their insulating oil, which serves both as a dielectric medium and a cooling agent. One of the most critical factors affecting transformer oil performance is moisture contamination, which significantly reduces dielectric strength, accelerates aging processes, and increases the probability of insulation breakdown. The purpose of this study is to develop and analyze a reliable method for monitoring the water content in transformer oil. A capacitive sensor system was designed and applied to measure variations in dielectric permittivity caused by different levels of moisture concentration. Experimental measurements were carried out using transformer oil samples with controlled humidity levels, and the obtained results demonstrated a clear correlation between the sensor output and the actual water content. The proposed method provides a simple, accurate, and real-time monitoring approach, which can be effectively used for predictive maintenance and ensuring the safe operation of power transformers.
-
Article: 13Published: 31.08.2025 pp. 43-54
Thermal, Exergy, and Environmental Performance Analysis of Modified Fin Geometries in Panel Radiators: A CFD-Based Study
- 369Views 279PDF access
AbstractIn this study, alternative fin geometries were designed and evaluated to enhance heat transfer performance in panel radiators. Computational Fluid Dynamics (CFD) simulations were performed using ANSYS software for the traditional design and three newly proposed fin profiles, providing detailed data on temperature distribution and thermal efficiency. Additionally, exergy efficiencies and annual CO₂ emissions were calculated for each configuration. Two of the newly designed fins exhibited thermal efficiencies close to the conventional fin (approximately 48.3%), while the third design—an I-profile fin—achieved the highest efficiency at 61.9%. The exergy efficiency increased from 13.3% in the conventional design to 15.8% in Fin 3, indicating a 2.5% improvement. Although the enhanced heat transfer resulted in a 12.5 kg increase in annual CO₂ emissions compared to the conventional design, this trade-off is considered acceptable when evaluated alongside the significant gains in energy and exergy efficiency. Overall, the findings suggest that with optimized fin geometries, it is possible to achieve a more sustainable radiator performance through a balanced approach between thermal efficiency and environmental impact.
