cover
Contact Name
Alfian Maarif
Contact Email
alfianmaarif@ee.uad.ac.id
Phone
-
Journal Mail Official
biste@ee.uad.ac.id
Editorial Address
-
Location
Kota yogyakarta,
Daerah istimewa yogyakarta
INDONESIA
Buletin Ilmiah Sarjana Teknik Elektro
ISSN : 26857936     EISSN : 26859572     DOI : 10.12928
Core Subject : Engineering,
Buletin Ilmiah Sarjana Teknik Elektro (BISTE) adalah jurnal terbuka dan merupakan jurnal nasional yang dikelola oleh Program Studi Teknik Elektro, Fakultas Teknologi Industri, Universitas Ahmad Dahlan. BISTE merupakan Jurnal yang diperuntukkan untuk mahasiswa sarjana Teknik Elektro. Ruang lingkup yang diterima adalah bidang teknik elektro dengan konsentrasi Otomasi Industri meliputi Internet of Things (IoT), PLC, Scada, DCS, Sistem Kendali, Robotika, Kecerdasan Buatan, Pengolahan Sinyal, Pengolahan Citra, Mikrokontroller, Sistem Embedded, Sistem Tenaga Listrik, dan Power Elektronik. Jurnal ini bertujuan untuk menerbitkan penelitian mahasiswa dan berkontribusi dalam pengembangan ilmu pengetahuan dan teknologi.
Arjuna Subject : -
Articles 295 Documents
Design and Implementation of PID Controller for Phase Shifted Full Bridge DC-DC Converter Ibrahim, Luay G.; Shneen, Salam Waley
Buletin Ilmiah Sarjana Teknik Elektro Vol. 7 No. 3 (2025): September
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/biste.v7i3.13814

Abstract

The study aims to evaluate the converter's operation and how to improve it, and to discuss the performance and behavior of the system in transient and steady states. Phase shifted full bridge DC-DC Converter (PSFB), converter is one of the most popular isolated converter, which is widely used in many applications. So today we are discussing its working, design and MATLAB simulation. converters are used frequently to step down high dc voltages to lower voltages. It also provides isolation between input and output stages. it is major application includes, server power supply, telecom rectifier, battery charging system and renewable energy systems. This is the basic structure of a full bridge DC-DC Converter. we have a dc voltage source at the input. there are the power electronic switches, which can be either MOSFET or IGBT. this is a high frequency transformer. used for the isolation between input and output stages. it also provides the required voltage gain. an inductor is used to limit the output current ripple, and a capacitor used as the filter to regulate the output voltage. To meet the necessary demand, in addition to regulating the quality of electrical power to address the changes and fluctuations in the system caused by various factors, the output of the converters is enhanced by developing a model design through simulation to provide the appropriate voltage, current, and power to cover the required load. Industrial applications are among the most important industries that employ and use electronic power converters, including the DC-DC converter, especially the PSFB. Among these applications are charging systems for storage units in electricity generation systems from renewable energies, including solar or wind energy, with a DC generator. It can also be part of a lighting system or microgrids, as this converter is characterized by high efficiency in performance, quality, and reliability, and has the advantage of a wide range at high frequencies. The PSFB converter consists of a DC source to supply a DC load, connected to an inverter on the source side and a filter on the load side, with a rectifier between them. The rectifier is a bridge type of four diodes, the inverter is a bridge of four MOSFET transistors, and the LC filter consists of a coil and a capacitor. Among the areas that require a wide frequency range are communications systems, which is one of the most important applications and areas of use for this converter.
Analysis of Improve Performance and Dynamics of an Induction Motor using an Artificial Neural Network Controller and a Conventional Proportional Integral Derivative Controller Shuraiji, Ahlam Luaibi; Shneen, Salam Waley
Buletin Ilmiah Sarjana Teknik Elektro Vol. 7 No. 3 (2025): September
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/biste.v7i3.13820

Abstract

Systems vary depending on the changing operating conditions. Some include linear systems, which previous studies have proven can be controlled using conventional systems, while non-linear systems require expert and intelligent controllers. To verify this, the current study compares expert artificial neural networks (ANNs) with traditional PID controllers for controlling the rotational speed of an induction motor. Traditional PID controllers are simple and easy to implement, but they lack the ability to handle changing operating conditions and do not have the capacity to adapt to load fluctuations as expert systems such as neural networks do. They also have the ability to handle load disturbances and are considered more effective, efficient, and robust compared to traditional PID controllers. PID controllers are easy to adjust and simple in structure, and are widely used with linear industrial systems. PID controllers have degraded performance when the load changes, i.e., when the system is non-linear, their performance deteriorates. ANN, on the other hand, are characterized by their ability to adapt to varying conditions and changing loads. In non-linear systems, they have the ability to adapt and handle system disturbances. ANNs are expensive and require precise design, data for network architecture, and training. The feasibility of tracking induction motor speed is investigated using motor simulation models, conventional PID controllers, and expert neural networks, and the simulation results are analyzed and compared. The simulation results demonstrate that ANNs outperform PIDs in response speed and lower overshoot and undershoot limits under various operating conditions. From the above, it can be concluded that expert neural networks can effectively control and improve dynamic response of induction motors due to their adaptive and learning capabilities, and they can handle nonlinear systems such as changing load conditions. It is proposed to conduct simulation tests of an electric motor using MATLAB engineering software, by mathematically representing it using a transfer function according to characteristics suitable for applications similar to the proposed characteristics. Simulation tests are conducted for an open circuit system, a closed circuit system without control, and a closed circuit system with control. The second method involves self-tuning the conventional controller to achieve the best design by optimizing performance, response speed, overshoot rate, and rise time, according to the proposed operating algorithm. The results demonstrate the superiority of the neural network over conventional controllers.
Transforming EEG into Scalable Neurotechnology: Advances, Frontiers, and Future Directions Pamungkas, Yuri; Triandini, Evi; Forca, Adrian Jaleco; Sangsawang, Thosporn; Karim, Abdul
Buletin Ilmiah Sarjana Teknik Elektro Vol. 7 No. 3 (2025): September
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/biste.v7i3.13824

Abstract

Electroencephalography (EEG) is a key neurotechnology that enables non-invasive, high-temporal resolution monitoring of brain activity. This review examines recent advancements in EEG-based neuroscience from 2021 to 2025, with a focus on applications in neurodegenerative disease diagnosis, cognitive assessment, emotion recognition, and brain-computer interface (BCI) development. Twenty peer-reviewed studies were selected using predefined inclusion criteria, emphasizing the use of machine learning on EEG data. Each study was assessed based on EEG settings, feature extraction, classification models, and outcomes. Emerging trends show increased adoption of advanced computational techniques such as deep learning, capsule networks, and explainable AI for tasks like seizure prediction and psychiatric classification. Applications have expanded to real-world domains including neuromarketing, emotion-aware architecture, and driver alertness systems. However, methodological inconsistencies (ranging from varied preprocessing protocols to inconsistent performance metrics) pose significant challenges to reproducibility and real-world deployment. Technical limitations such as inter-subject variability, low spatial resolution, and artifact contamination were found to negatively impact model accuracy and generalizability. Moreover, most studies lacked transparency regarding bias mitigation, dataset diversity, and ethical safeguards such as data privacy and model interpretability. Future EEG research must integrate multimodal data (e.g., EEG-fNIRS), embrace real-time edge processing, adopt federated learning frameworks, and prioritize personalized, explainable models. Greater emphasis on reproducibility and ethical standards is essential for the clinical translation of EEG-based technologies. This review highlights EEG’s expanding role in neuroscience and emphasizes the need for rigorous, ethically grounded innovation.
Implementation of Filter Bank Multicarrier Transmitter Using Universal Software Radio Peripheral Abdulhussein, Ali A.; Abdullah, Hikmat N.; Marhoon, Hamzah M.
Buletin Ilmiah Sarjana Teknik Elektro Vol. 7 No. 3 (2025): September
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/biste.v7i3.13876

Abstract

Particularly in 5G and beyond, Filter Bank Multicarrier (FBMC) modulation is becoming more widely acknowledged as a potent substitute for traditional Orthogonal Frequency Division Multiplexing (OFDM) in upcoming wireless communication systems. FBMC is robust in situations impacted by multipath fading, synchronisation errors, and spectral leakage because of its improved spectral efficiency, superior time-frequency localisation, and removal of cyclic prefix. The Universal Software Radio Peripheral (USRP) N210, combined with GNU Radio version 3.,7 is used in this paper to design and implement a working FBMC transmitter. The system architecture supports 32 and 64 subcarrier allocations that can be changed to accommodate different communication scenarios, allowing for real-time signal generation and transmission. While SDR hardware was used for transmission and reception, software was used to develop the entire signal processing chain, including modulation, prototype filtering, and transmission. To evaluate performance metrics like constellation accuracy, spectrum containment, and signal quality, a number of experiments were carried out. These tests validate the viability of the suggested SDR-based architecture in real-world settings by confirming the successful generation and over-the-air transmission of FBMC signals. Notably, the work tackles important real-time implementation issues like subcarrier reconfigurability, synchronisation overhead, and hardware limitations. Along with its usefulness, this implementation lays the groundwork for future improvements by incorporating clever optimization algorithms like Harris Hawks Optimization, Particle Swarm Optimization, and Genetic Algorithms. Aspects like filter design, subcarrier spacing, and power efficiency can all be enhanced by utilising these algorithms. According to the results, the suggested system is in a good position to be implemented in adaptive and cognitive radio applications, where resilience to changing channel conditions and effective spectrum utilization is essential.
A Generalized Deep Learning Approach for Multi Braille Character (MBC) Recognition Widyadara, Made Ayu Dusea; Handayani, Anik Nur; Herwanto, Heru Wahyu; Yu, Tony
Buletin Ilmiah Sarjana Teknik Elektro Vol. 7 No. 3 (2025): September
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/biste.v7i3.13891

Abstract

Automated visual recognition of Multi Braille Characters (MBC) poses significant challenges for assistive reading technologies for the visually impaired. The intricate dot configurations and compact layouts of Braille complicate MBC classification. This study introduces a deep learning approach utilizing Convolutional Neural Networks (CNN) and compares four leading architectures: ResNet-50, ResNet-101, MobileNetV2, and VGG-16. A dataset comprising 105 MBC classes was developed from printed Braille materials and underwent preprocessing that included image cropping, brightness enhancement, character position labeling, and resizing to 89×89 pixels. A 70:20:10 data partitioning strategy was applied for training and evaluation, with variations in batch sizes (8–128) and epochs (50–500). The results demonstrate that ResNet-101 achieved superior performance, attaining an accuracy of 91.46%, an F1-score of 89.48%, and a minimum error rate of 8.5%. ResNet-50 and MobileNetV2 performed competitively under specific conditions, whereas VGG-16 consistently exhibited lower accuracy and training stability. Standard deviation assessments corroborated the stability of residual architectures throughout the training process. These results endorse ResNet-101 as the most effective architecture for Multi Braille Character classification, highlighting its potential for incorporation into automated Braille reading systems, a tool for translating braille into text or sound for future needs.
Analysis of Swarm Size and Iteration Count in Particle Swarm Optimization for Convolutional Neural Network Hyperparameter Optimization in Short-Term Load Forecasting Nguyen, Tuan Anh; Nguyen, Trung Dung
Buletin Ilmiah Sarjana Teknik Elektro Vol. 7 No. 3 (2025): September
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/biste.v7i3.13953

Abstract

Short-term load forecasting (STLF) is critical in modern power system planning and operation. However, the effectiveness of deep learning models such as Convolutional Neural Networks (CNNs) depends on selecting hyperparameters, which are traditionally tuned through time-consuming trial-and-error processes. The research contribution of this study is to systematically analyze how two key parameters—swarm size and iteration count—in Particle Swarm Optimization (PSO) affect the performance of CNN hyperparameter tuning for STLF. A CNN architecture with fixed convolutional depth is optimized using PSO over selected hyperparameters, including the number of filters, batch size, and training epochs. The experiments use two regional Australian electricity load datasets: New South Wales (NSW) and Queensland (QLD). A three-fold cross-validation strategy is employed, and the Mean Absolute Percentage Error (MAPE) is used as the primary evaluation metric. The results show that optimal PSO configurations vary significantly between datasets, with smaller swarm sizes and moderate iteration counts yielding favorable trade-offs between forecasting accuracy and computational cost. However, the reliance on MAPE, sensitivity to near-zero values, and fixed CNN architecture impose limitations. This study provides practical guidance for selecting PSO settings in deep learning-based STLF and demonstrates that tuning PSO configurations can significantly enhance model performance while reducing computational overhead. Future work may explore adaptive or hybrid optimization methods and extend to more diverse forecasting scenarios.
Resource-Efficient Sentiment Classification of App Reviews Using a CNN-BiLSTM Hybrid Model Baktibayev, Daulet; Serek, Azamat; Berlikozha, Bauyrzhan; Rustauletov, Babur
Buletin Ilmiah Sarjana Teknik Elektro Vol. 7 No. 3 (2025): September
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/biste.v7i3.13954

Abstract

This study evaluates the performance of a hybrid convolutional neural network and bidirectional long short-term memory (CNN + BiLSTM) model for sentiment classification on user reviews from the Spotify mobile application. The primary aim is to explore whether competitive results can be achieved without relying on transformer-based architectures, which often require substantial computational resources. The proposed CNN + BiLSTM model combines local feature extraction with sequential context modeling and is benchmarked against traditional machine learning and simpler deep learning models, including a Random Forest classifier enhanced with polarity features, a standalone CNN, and a fully connected DNN. Sentiment labels were binary (positive or negative) and directly provided in the dataset without being inferred from star ratings. The dataset was balanced to avoid class skew. Experimental results indicate that the CNN + BiLSTM model achieves moderate improvements over the baseline models, with an accuracy of 0.8861 and an F1-score of 0.8691. While it does not surpass the highest-performing transformer-based methods reported in the literature, it performs comparably to several of them, despite having a lower computational footprint. Analyses of ROC curves, confusion matrices, and training dynamics further contextualize the model’s performance, showing strengths in classifying negative sentiments and convergence efficiency. To address overfitting, early stopping and dropout layers were employed as regularization techniques. The study contributes to the ongoing discourse on resource-efficient sentiment analysis by showing that hybrid architectures may offer a practical balance between model complexity and performance in specific application domains.
Small Object Detection in Medical Imaging Using Enhanced CNN Architectures for Early Disease Screening Zangana, Hewa Majeed; Omar, Marwan; Li, Shuai; Al-Karaki, Jamal N.; Vitianingsih, Anik Vega
Buletin Ilmiah Sarjana Teknik Elektro Vol. 7 No. 3 (2025): September
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/biste.v7i3.14015

Abstract

Early detection of subtle pathological features in medical images is critical for improving patient outcomes but remains challenging due to low contrast, small lesion size, and limited annotated data. The research contribution is a hybrid attention-enhanced CNN specifically tailored for small object detection across mammography, CT, and retinal fundus images. Our method integrates a ResNet-50 backbone with a modified Feature Pyramid Network, dilated convolutions for contextual scale expansion, and combined channel–spatial attention modules to preserve and amplify fine-grained features. We evaluate the model on public benchmarks (DDSM, LUNA16, IDRiD) using standardized preprocessing, extensive augmentation, and cross-validated training. Results show consistent gains in detection and localization: ECNN achieves an F1-score of 88.2% (95% CI: 87.4–89.0), mAP@0.5 of 86.8%, IoU of 78.6%, and a low false positives per image (FPPI = 0.12) versus baseline detectors. Ablation studies confirm the individual contributions of dilated convolutions, attention modules, and multi-scale fusion.However, these gains involve higher computational costs (≈2× training time and increased memory footprint), and limited dataset diversity suggests caution regarding generalizability. In conclusion, the proposed ECNN advances small-object sensitivity for early disease screening while highlighting the need for broader clinical validation and interpretability tools before deployment.
Transfer Learning Models for Precision Medicine: A Review of Current Applications Pamungkas, Yuri; Aung, Myo Min; Yulan, Gao; Uda, Muhammad Nur Afnan; Hashim, Uda
Buletin Ilmiah Sarjana Teknik Elektro Vol. 7 No. 3 (2025): September
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/biste.v7i3.14286

Abstract

In recent years, Transfer Learning (TL) models have demonstrated significant promise in advancing precision medicine by enabling the application of machine learning techniques to medical data with limited labeled information. TL overcomes the challenge of acquiring large, labeled datasets, which is often a limitation in medical fields. By leveraging knowledge from pre-trained models, TL offers a solution to improve diagnostic accuracy and decision-making processes in various healthcare domains, including medical imaging, disease classification, and genomics. The research contribution of this review is to systematically examine the current applications of TL models in precision medicine, providing insights into how these models have been successfully implemented to improve patient outcomes across different medical specialties. In this review, studies sourced from the Scopus database, all published in 2024 and selected for their "open access" availability, were analyzed. The research methods involved using TL techniques like fine-tuning, feature-based learning, and model-based transfer learning on diverse datasets. The results of the studies demonstrated that TL models significantly enhanced the accuracy of medical diagnoses, particularly in areas such as brain tumor detection, diabetic retinopathy, and COVID-19 detection. Furthermore, these models facilitated the classification of rare diseases, offering valuable contributions to personalized medicine. In conclusion, Transfer Learning has the potential to revolutionize precision medicine by providing cost-effective and scalable solutions for improving diagnostic capabilities and treatment personalization. The continued development and integration of TL models in clinical practice promise to further enhance the quality of patient care.
Design and Application of a Cyber Physical Based Data Logger System for Charging Stations Rahutomo, Faisal; Nugraha, Bagus Putra; Mekonnen, Atinkut Molla; Ariyo, Bashiru Olalekan
Buletin Ilmiah Sarjana Teknik Elektro Vol. 7 No. 3 (2025): September
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/biste.v7i3.13266

Abstract

The rapid advancement of technology, particularly in transportation, has led to a growing public interest in electric vehicles. Government support, exemplified by Presidential Regulation No. 55 of 2019, further encourages this shift. With more electric vehicles on the road, the need for adequate charging infrastructure is critical. This research aims to design, test, and implement a charging device that records electric vehicle usage, displays data on an LCD, and allows monitoring through a website. Using the research and development (R&D) method, a highly effective design was developed. The data recording system employs the PZEM-004T sensor and ESP32 microcontroller to send data to a database. Validation tests showed high accuracy and precision, with current accuracy at 98.79% and precision at 99.24%, and voltage accuracy at 99.59% and precision at 99.87%. The device was installed in the basement of UPT TIK UNS and tested with three electric vehicles, each with different power requirements. The average power growth every ten minutes was 0.063 kWh for the first vehicle, 0.164 kWh for the second, and 0.139 kWh for the third. These results demonstrate that the device functions well, the design is successful, and it provides consistent, accurate, and precise energy growth measurements.