Jurnal Nasional Teknik Elektro
Not a member yet
359 research outputs found
Sort by
Minimizing THD Using a Multilevel Inverter Integrated with MPPT
Full text linkThis paper presents a novel Modified Multilevel Inverter (MMLI) topology to reduce Total Harmonic Distortion (THD) in photovoltaic (PV) systems. Unlike conventional Cascaded H-Bridge Inverters, the proposed MMLI achieves higher output voltage levels using fewer switching components by optimizing the arrangement of voltage sources and switches. A Boost converter integrated with Maximum Power Point Tracking (MPPT) further enhances power conversion efficiency and system stability. This specific configuration has not been previously explored, offering a more effective solution for THD mitigation. Simulations conducted in MATLAB/Simulink demonstrate the inverter’s performance, with the 5-level MMLI achieving a THD of 28.99% and the 9-level configuration reducing it to 18.15%. These results confirm the superiority of the proposed topology in improving power quality and reducing system complexity. Moreover, the design eliminates the need for external filters, making it a cost-effective and practical option for grid-connected PV applications
Incubator for Joper Day Old Chicks with Cohen-Coon PID Controller
Full text linkJoper (Jowo Super) chicken parents are generally unable to incubate eggs or provide adequate care for their offspring, making the use of a specialized incubator essential from the hatching phase to Day-Old Chick (DOC) rearing. One of the primary factors contributing to DOC mortality is improper temperature adjustment in the heating system. To ensure optimal early-age development, Joper DOC requires a stable thermal environment within the range of 32°C to 35°C, depending on the growth stage. This study aims to develop an incubator capable of maintaining a constant temperature of 32 °C using Cohen-Coon PID (C-C PID) control while also regulating humidity levels. The proposed incubator integrates an axial fan and an L298N driver, with the temperature and humidity sensors calibrated prior to use. The calibration results show measurement errors of 0.59% for temperature and 5.02% for humidity, indicating high reliability. The application of C-C PID control demonstrates strong performance, characterized by a short rise time (approximately 225 seconds), an acceptable settling time (around 510 seconds), minimal overshoot (1.56%), and a steady-state error approaching 0%. During a 30-minute evaluation period, the system successfully maintained a stable temperature at the 32 °C set point and controlled humidity at below 50% automatically. Furthermore, the incubator design proved effective in practical use, achieving a 0% mortality rate for Joper DOC
A Hybrid Wavelet Scattering and Mel Spectrogram Feature with Deep Convolution Neural Network for Robust Spoken Digit Recognition
Full text linkSpoken digit recognition (SDR) plays a critical role in biometric authentication and human–computer interaction, yet existing approaches often rely on small datasets, limited feature representations, or architectures prone to overfitting. To address these limitations, this study proposes a robust end-to-end pipeline that integrates Wavelet Time Scattering (WTS), Mel-Frequency Cepstral Coefficients (MFCC), and a 2D Deep Convolutional Neural Network (2D-CNN) to enhance the accuracy and generalization of SDR systems in realistic environments. The Free-Spoken Digit Dataset (FSDD), consisting of 3000 audio samples from speakers with diverse accents, was pre-processed using zero-padding normalization and transformed into high-resolution time–frequency spectrograms via WTS. The proposed CNN architecture, optimized through systematic experimentation on batch size and learning rate, demonstrated stable convergence and superior discriminative capability. Using a learning rate of 0.001 and a batch size of 50, the model achieved the highest performance with 99.2% accuracy, outperforming established methods including SVM, MFCC-LSTM, and Multiple RNN architectures. Comparative evaluations further revealed that the combined WTS–MFCC feature extraction significantly enhances spectral–temporal representation quality, contributing to improved classification precision across all digit classes. These findings demonstrate that the proposed WTS-MFCC-CNN framework not only advances SDR accuracy but also provides a scalable and computationally efficient approach suitable for real-world biometric, financial, and voice-controlled applications. The results highlight the potential of hybrid time–frequency representations integrated with deep architectures to set a new benchmark for robust spoken digit recognition
Deteksi biji kakao kering dan basah menggunakan Convolutional Neural Network
No full textCocoa is one of Indonesia’s leading export commodities crucial in supporting the national economy. However, a significant challenge in the post-harvest processing of cocoa lies in the drying stage, where uneven drying often leads to inconsistent bean quality. While previous studies have predominantly focused on classifying cocoa beans based on surface defects such as faded, non-faded, fragmented, moldy, and damaged beans, limited research has addressed classification based on moisture levels—specifically distinguishing between dry and wet beans, which is essential for ensuring optimal fermentation, proper storage, and overall product quality. This study presents a classification model based on a Convolutional Neural Network (CNN) employing the You Only Look Once (YOLO) architecture to detect and classify dry and wet cocoa beans by analyzing visual features, particularly color and shape. A dataset of 2,880 labeled images was compiled and used to train and evaluate the model. The proposed system demonstrated strong performance, achieving an accuracy of 99.8%, a precision of 99.15%, and a recall of 99.8%. These results indicate that the model can be a reliable and efficient tool for detecting moisture content in cocoa beans, thereby enhancing quality control, reducing human subjectivity in the inspection process, and supporting automation in the cocoa processing industry to ensure product consistency and added value in the export market
Evaluasi Pengaruh Tekanan-Arus pada Kehilangan Fiber melalui NIRS DA1650
No full textThis study focuses on enhancing the yield of crude palm oil (CPO) during the pressing process by thoroughly examining the oil losses that occur throughout production. The primary aim is to evaluate how different pressures and electric currents impact oil losses from palm fiber at a specific palm oil mill in Pantai Cermin, Kec. Tapung, Kampar, Riau. A systematic methodology was employed to achieve this, which involved detailed measurements conducted using the FOSS NIRS DA1650. This advanced technology allowed for precise assessment and quantification of oil losses during the pressing phase. Following the data collection, a rigorous statistical analysis was performed utilizing determination coefficients to interpret the relationship between the variables. The analysis results revealed a coefficient of determination (R²) of 49.96% concerning pressure, suggesting that nearly half of the variability in oil losses can be explained by fluctuations in pressing pressure. Additionally, the examination of current showed a higher coefficient of determination of 60.09%, underscoring a substantial influence of electric current on fiber oil losses. These findings highlight the critical importance of optimizing pressure and current in palm oil extraction. By making informed adjustments to these parameters, mill operators can significantly reduce oil losses, thus enhancing the overall extraction efficiency. The study provides practical recommendations for operators aiming to improve their processes, ultimately contributing to better resource utilization and increased profitability in the palm oil industry.Industri pengolahan kelapa sawit menghadapi masalah utama dalam mengoptimalkan hasil CPO dengan fokus pada pengurangan kehilangan minyak selama proses produksi. Salah satu tantangan signifikan dalam mencapai tujuan ini adalah tingginya tingkat kehilangan minyak selama ekstraksi, terutama pada mesin pemeras serat, yang memerlukan kondisi optimal. Penelitian ini bertujuan untuk menganalisis kehilangan minyak pada unit screw press di pabrik kelapa sawit yang terletak di Pantai Cermin, Kec. Tapung, Kampar, Riau, serta mengkaji pengaruh tekanan dan arus listrik terhadap kehilangan minyak pada serat. Kehilangan minyak diukur menggunakan FOSS NIRS DA1650 selama sepuluh kali percobaan, dengan nilai berkisar antara 3,80% hingga 5,83%. Analisis korelasi menunjukkan bahwa tekanan memiliki koefisien korelasi Pearson (r) sebesar -0,70683, yang menunjukkan hubungan terbalik yang kuat dengan kehilangan minyak dan koefisien determinasi sebesar 49,96%. Hal ini menunjukkan bahwa tekanan mempengaruhi kehilangan minyak, sementara 50,04% dipengaruhi oleh faktor lain. Arus listrik menunjukkan koefisien korelasi Pearson (r) sebesar -0,77518 dan koefisien determinasi sebesar 60,090%, menunjukkan pengaruh kuat terhadap kehilangan minyak pada serat, dengan sisa 39,91% dipengaruhi oleh faktor lain
Stock of Nuts/Bolts System With A Load Cell Sensor of Digital Scale As An Iot-Based
No full textThe passage highlights the importance of having a reliable supply of nuts and bolts for Micro, Small, and Medium Enterprises (MSMEs) involved in the manufacturing of appropriate technology tools. Nuts and bolts are critical components in the production process of such manufacturing. Therefore, it is essential to ensure their constant availability to avoid interruptions in the workflow. Currently, many MSMEs still monitor their stock levels manually. This manual method has its drawbacks, as stock levels can sometimes go unnoticed, leading to stockouts when the components are needed. When stock runs out unexpectedly, it can cause delays, as additional time is required to purchase more supplies from wholesalers. In this study, a digital scale with a load cell sensor was developed as a medium for taking inventory or stocking bolts and nuts. This device features several buttons: a button for selecting the size of bolts or nuts, a button for choosing between stocking or taking items, and a button for displaying stock information of bolts or nuts in the warehouse. The results from this digital scale are sent to a website using the Internet of Things (IoT) system as a communication medium between the digital scale and the monitoring website. The results of this study show that the digital scale has an accuracy of 99.95%, and the accuracy of the counted items or stock is 100%.CV Baja Diva yang berada di jalan Lintas Timur KM 17, Kulim Pekanbaru Riau merupakan usaha mikro dalam bidang manufaktur dalam pembuatan peralatan teknologi tepat guna. Baut dan mur merupakan salah satu komponen yang sangat di butuhkan dalam memproduksi pembuatan manufaktur teknologi tepat guna di CV Baja Diva. Sementara saat ini untuk stok barang baut dan mur di CV Baja Diva masih menggunakan system manual. Pada penilitian ini dibuat timbangan digital dengan sensor loadcell sebagai media pengambilan atau stok barang mur dan baut. Adapun alat ini terdapat beberapa tombol yaitu tombol pilihan jenis ukuran baut atau mur, tombol pilihan stok barang atau ambil barang, tombol informasi stok mur atau baut di gudang. Hasil dari timbangan digital ini di kirimkan ke web menggunakan system Internet of Thing (IoT) sebagai sarana komunikasi antara timbangan digital dengan monitoring di Website. Adapun hasil penelitian ini memiliki akurasi timbangan digital sebesar 99,95 % dan akurasi jumlah barang yang terambil atau stok barang sebesar 100%
Development of DC Motor Speed Control Using PID Based on Arduino and Matlab For Laboratory Trainer
Full text linkDC motors are widely used as propulsions, including in electric bicycles. The problem faced by students in the DC motor control laboratory working using software simulation is that they do not have practical learning experience using digital instruments. This article aims to develop a DC motor speed control that can be used to learn practical Proportional Integral Derivative (PID) control in the laboratory. The DC motor speed control was developed using a combination of Arduino UNO microcontroller and Matlab software. The PID method was used because it is still broadly studied and applied in industries. The test results showed that the developed trainer can work well with PID variable values that can be entered via the keypad, and DC motor transient responses can be displayed in Matlab. From the experimental results, it was found that the optimal PID variable values were Kp=0.04, Ki=0.05, and Kd=0.004, where the controller produced a low overshoot value, i.e., 0.73% of its set point and a settling time of 10.66 seconds. The test results of using the developed trainer in the Fundamental of Control Engineering laboratory work showed that the developed trainer gave students practical learning experience in designing PID control for DC motor speed by using digital equipment, i.e., microcontroller and actual DC motor as well as analyzing its corresponding transient response in Matlab software environmen
Pengembangan IDS pada IoT Menggunakan Ensemble Learning
No full textThe utilization of intrusion detection systems (IDS) can significantly enhance the security of IT infrastructure. Machine learning (ML) methods have emerged as a promising approach to improving the capabilities of IDS. The primary objective of an IDS is to detect various types of malicious intrusions with a high detection rate while minimizing false alarms, surpassing the capabilities of a firewall. However, developing an IDS for IOT poses substantial challenges due to the massive volume of data that needs to be processed. To address this, an optimal approach is required to improve the accuracy of data containing numerous attacks. In this study, we propose a novel IDS model that employs the Random Forest, Decision Tree, and Logistic Regression algorithms using a specialized ML technique known as Ensemble Learning. For this research, we used the BoT-IoT datasets as inputs for the IDS model to distinguish between malicious and benign network traffic. To determine the best model, we compared the performance metrics of each algorithm across different parameter combinations. The research findings demonstrate exceptional performance, with metric scores exceeding 99.995% for all parameter combinations. Based on these conclusive results, we deduce that the proposed model achieves remarkable success and outperforms other traditional ML-based IDS models in terms of performance metrics. These outcomes highlight the potential of our novel IDS model to enhance the security posture of IoT-based systems significantly.Pemanfaatan sistem deteksi intrusi (IDS) memiliki potensi untuk meningkatkan postur keamanan infrastruktur TI. IDS berfungsi sebagai metode yang penting dan praktis untuk mendeteksi serangan dan menjaga keamanan jaringan dari peretas yang mengganggu. Metode Machine Learning (ML) telah muncul sebagai pendekatan yang menjanjikan untuk meningkatkan kemampuan IDS. Ada banyak model klasifikasi yang telah dikembangkan untuk membangun sistem IDS yang efisien, yang memainkan peran penting dalam melindungi sistem jaringan. Namun, pemanfaatan algoritma pembelajaran ensemble masih relatif terbatas. Pendekatan yang optimal diperlukan untuk meningkatkan akurasi data yang mengandung banyak serangan. Pada penelitian ini, model IDS baru menggunakan algoritma Random Forest, Decision Tree, dan Logistic Regression, yang merupakan teknik ML khusus menggunakan Ensemble Learning. Metodologi penelitian melibatkan dataset BoT-IoT, yang berfungsi sebagai input untuk model IDS untuk membedakan antara lalu lintas jaringan yang jinak dan berbahaya. Untuk mengidentifikasi model terbaik, metrik kinerja dari masing-masing model dibandingkan dengan kombinasi parameter yang berbeda. Selanjutnya, model akhir menjalani evaluasi menggunakan teknik pembelajaran ensemble untuk memastikan kinerja yang optimal. Temuan penelitian menunjukkan skor metrik kinerja melebihi 99,995% untuk semua kombinasi parameter. Berdasarkan hasil konklusif ini, kami menyimpulkan bahwa model yang diusulkan tidak hanya mencapai kesuksesan tetapi juga mengungguli model IDS berbasis ML tradisional lainnya dalam hal metrik kinerja
Comparative Analysis of Two-Stage and Single-Stage Models in Batteryless PV Systems for Motor Power Supply
Full text linkImplementing photovoltaic (PV) systems as direct power sources for motors without batteries is a complex process that requires a sophisticated control mechanism. The crucial aspect of PV systems is the Maximum Power Point Tracking (MPPT) process, which ensures that the installed PV system generates optimal energy output. A recent study has analyzed research related to PV systems supplying power to pump motors, and the results have successfully classified these systems into two main models: the two-stage and the single-stage. The two-stage model involves separate power tracking and load consumption control processes, while the single-stage model integrates power tracking and load consumption control into a single process. A comparative analysis of these two models has revealed that the two-stage model exhibits higher stability due to the separate power tracking and load consumption control processes. Aspects such as the MPPT process, motor power consumption, and the utilization of DC-link capacitors were examined in this study. The findings of this comparative study contribute valuable insights into the effectiveness and stability of two-stage and single-stage models in PV systems supplying power to motors without batteries. The results will significantly interest researchers and practitioners working in Photovoltaic systems and motor control, providing helpful information for designing and implementing more efficient and reliable PV systems
A Techno-Economic Analysis for Raja Ampat Off-Grid System
Full text linkIndonesia, an expansive archipelagic nation with over 17,000 islands, encounters significant challenges in ensuring a sustainable and dependable electricity supply, particularly in its West Papua region. The reliance on diesel fuel for electricity generation in this area poses substantial environmental risks and incurs high costs. A comprehensive research study addressing the environmental and economic challenges associated with diesel dependence in West Papua proposed a shift towards sustainable and cost-effective solutions by advocating for adopting off-grid hybrid power systems. This study targeted Yensawai Village in the Raja Ampat Islands, employing a detailed techno-economic analysis through HOMER Pro to identify the most cost-effective system configurations. The findings indicated that the optimal setup consists of a 160 kW diesel generator, complemented by a 70.1 kW solar photovoltaic (PV) system, a 30 kW inverter, and an 80 kWh battery storage unit. This configuration not only proved to be economically viable by reducing the levelized cost of electricity (CoE) by 15.7%—achieving a CoE of 0.280/kWh—but also highlighted the potential for similar benefits across regional systems. By focusing on the economic advantages of hybrid energy configurations, this research contributes significantly to the broader discourse on sustainability and the urgent need to reduce diesel dependence, offering a practical approach to cutting electricity generation costs in remote island communities and advancing sustainability initiatives