TELKOMNIKA (Telecommunication Computing Electronics and Control)
Not a member yet
3120 research outputs found
Sort by
Machine learning-based reconstruction of missing rainfall extremes: a comparative analysis with classical models
Full text linkThe limited availability of daily rainfall data remains a key challenge in rainfall data analysis. This study assesses the effectiveness of spatial interpolation and bias correction techniques using satellite-derived rainfall data to fill missing observations in the Banten and Jakarta regions. Three interpolation methods inverse distance weighting (IDW), kriging, and spline were compared. Nine statistical and machine learning-based bias correction methods were applied to climate hazards group infrared precipitation with station data (CHIRPS), multi-source weighted-ensemble precipitation (MSWEP), and global precipitation measurement-integrated multi-satellite retrievals for GPM (GPM IMERG). Performance was evaluated using root mean square error (RMSE), mean absolute error (MAE), bias, Pearson correlation (R), and Kling-Gupta efficiency (KGE) in the expert team on climate change detection and indices (ETCCDI) extreme index. The research findings indicate that CHIRPS with quantile mapping (QM) bias correction delivers the best performance, followed by random forest regression (RFR) as the most accurate machine learning method. In spatial interpolation, IDW stands out as the leading method. Testing the extreme index ETCCDI confirms that CHIRPS-QM consistently outperforms machine learning and interpolation methods. In general, CHIRPS-QM and IDW represent the most effective combination of techniques for reconstructing daily rainfall, particularly extreme events. This study uniquely integrates spatial interpolation and bias correction in a unified evaluation
Recognition and understanding of construction safety signs by final year engineering students
Full text linkThis study assessed the recognition and understanding of construction safety signs among final-year Higher National Diploma, Building Technology (BT) and Electrical/Electronic Engineering (EE) students at a technical university in Ghana. The purpose was to evaluate their awareness of safety signs, addressing gaps in existing research and providing updated data to enhance occupational safety training. A descriptive statistical methodology was employed, utilizing purposive sampling to survey 137 students via structured questionnaires. Data were analyzed using SPSS v16 and compared against ISO 3864 and ANSI Z5353 standards. Results revealed varying comprehension rates: prohibition signs (61.71%), general warning signs (71.08%), mandatory signs (78.32%), emergency escape signs (81.4%), firefighting signs (86.9%), and chemical labeling signs (77.98%). While mean scores exceeded benchmark thresholds, low response rates for specific signs indicated significant knowledge gaps. The study concludes that unfamiliarity with safety signs persists due to insufficient training and curricular emphasis. Recommendations include revising academic syllabi under Ghana Tertiary Education Commission and National Board for Technical Examinations guidelines to integrate safety education, alongside industry partnerships for practical training during internships. These measures aim to reduce workplace accidents and improve safety compliance among future engineers
Novel internet of things-spectroscopy methods for targeted water pollutants in household point-of-use environments
Full text linkEnsuring water quality remains a paramount concern to prevent adverse health effects on consumers. Water quality monitoring primarily focuses on water utilities and infrastructure, such as treatment plants and reservoirs. More information is needed on the status of water once it enters the consumption phase, particularly at the point-of-use (POU). Therefore, this study aims to provide a scientific understanding of water quality in response to microbial contaminants in Malaysia’s household water system using the non-invasive benchtop near-infrared (NIR)-Raman spectroscopy approach. This study also provided the effects of seasonal variations and stagnation periods on the quality of water supply, corresponding to microbial contaminants. Findings show that almost 20% of the water samples contained Legionella and Salmonella species through the Raman spectral identification technique. The distinct signature peaks (ranging from 400 cm-1 to 1,800 cm-1) indicative of specific bacterial species are identified. However, benchtop Raman spectroscopy has application constraints in real-time water quality monitoring. Hence, acknowledging its limitation, this study proposed a new internet of things (IoT)-based micro-spectrometer as an alternative to rapid and sustainable POUs water quality assessment. Leveraging IoT protocols enhances the reliability and efficiency of identifying microbiological threats in water supply
A 6G THz MIMO antenna with high gain and wide bandwidth for high-speed wireless communication
Full text linkThis study presents a comprehensive industrial and innovation design and thorough analysis of a terahertz (THz) multiple-input multiple-output (MIMO) antenna, addressing the increasing demand for high-performance multi-antenna systems in THz communication applications. The primary objective of this research is to develop a compact and efficient MIMO antenna that operates over a wide frequency range and provides high isolation, specifically within the 1–10 THz spectrum. The proposed antenna achieves an impressive total bandwidth of approximately 9 THz, featuring seven distinct resonance frequencies at 1.39 THz, 3.26 THz, 4.72 THz, 5.96 THz, 7.07 THz, 8.194 THz, and 9.426 THz. The design employs a polyimide substrate and a graphene patch. Key performance metrics include a maximum gain of 15 dB, efficiency of 99.8%, and isolation values that range from 28 dB to 63 dB. An resistor inductor capacitor (RLC) equivalent circuit using advanced design system (ADS) software. Additionally, the antenna displays remarkable diversity metrics, with an envelope correlation coefficient (ECC) of 0.000778 and a diversity gain of 9.99961 dB. With compact dimensions of (65×180) µm2 and outstanding performance characteristics, this design is confirmed to be suitable for THz applications, fulfilling the research goal of facilitating efficient and reliable communication in sophisticated multi-antenna systems
A multiband sub-6 THz patch antenna with high gain for IoT and 6G communication
Full text linkThis comprehensive study introduces a meticulously designed and characterized terahertz (THz) multiple-input multiple-output (MIMO) antenna engineered to operate within the 0.4 THz to 1.6 THz frequency range. The antenna’s construction includes a copper patch and ground plane integrated into a polyimide substrate, ensuring exceptional durability and robust performance. Significantly, the antenna reveals four distinct resonance frequencies at 0.46 THz, 0.9 THz, 1.31 THz, and 1.44 THz each accompanied by bandwidths of 0.005 THz, 0.17 THz, and 0.34 THz, respectively. Moreover, the antenna delivers notable gains of 8.52 dB, 11.54 dB, and 13.25 dB at these frequencies, coupled with substantial efficiencies of 88.32%, 92.02%, and 89.89%, respectively. Additionally, the antenna showcases exceptional isolation of 26 dB, a low envelope correlation coefficient (ECC) of 0.003, and a diversity gain (DG) of 9.98. These remarkable attributes underscore the antenna’s aptness for high-performance THz applications, offering substantial advantages in terms of gain, efficiency, and isolation for next-generation wireless communication systems
Development of hydraulic servo controller for mechanical testing with optimization of PID tuning methods
Full text linkThis study explores the use of hydraulic servo control (HSC) systems in static and dynamic structural testing, focusing on optimizing proportional, integral, derivative (PID) controller tuning. The HSC system comprises three main components: hydraulic, control, and measurement systems. To achieve optimal performance, the research begins with preparing setpoint displacement/force data and developing mathematical models for the cylinder actuator and servo valve, incorporating sensors like load cells and linear variable differential transducers (LVDTs). A closed-loop transfer function is used to predict outputs that align closely with setpoint values. Three PID tuning methods—Ziegler-Nichols, Cohen-Coon, and adaptive control—are evaluated. Simulation results show all methods yield satisfactory performance with evaluation errors below 1.5%. Implementation tests further confirm effectiveness, with root mean square deviation (RMSD) values under 1%, indicating high precision. Despite promising results, the study acknowledges limitations due to restricted datasets and test conditions. Future research should address broader dynamic load variations, nonlinearities such as fluid leakage and hysteresis, and integrate intelligent optimization techniques like machine learning to enhance robustness and adaptability. This work contributes to improving the reliability and accuracy of HSC systems in structural testing, paving the way for smarter, more responsive control strategies in engineering applications
Optimal active disturbance rejection control with applications in electric vehicles
Full text linkThis work proposes an optimal control strategy based on a modified active disturbance rejection control (ADRC) that considers disturbance weighting for a three-phase induction motor under rotor field-oriented control (FOC) to enhance energy efficiency. Induction motors (IMs) are widely used in electric vehicles (EVs) due to their cost-effectiveness and technological maturity. However, improving energy efficiency remains a key challenge, as it directly impacts vehicle range. The proposed approach employs ADRC, where part of the disturbance rejection task is handled offline by a hybrid optimization algorithm combining particle swarm optimization (PSO), tabu search (TS), and simulated annealing (SA) to tune a state-feedback controller. The controller parameters are optimized using a composite cost function that balances energy consumption and performance. Simulation and experimental results indicate that disturbance weighting has a significant impact on both problem complexity and performance. Optimal weighting improves the overall system response compared to conventional disturbance rejection methods. Energy and performance analyses show that disturbance weighting enhances energy usage compared to the traditional ADRC method, suggesting a novel efficiency control strategy for electric machines
A dual-band modified-rectangular patch with parasitic antenna for 2.4/5 GHz wireless local area network applications
Full text linkThis research presents the design and implementation of a dual-band patch antenna (DBPA) optimized for 2.4 GHz and 5 GHz wireless local area network (WLAN) applications. The antenna features a modified rectangular patch with a cut corner and two parasitic rectangular patches, enabling dual-band operation with enhanced gain. The DBPA is fed by a 50-Ohm coplanar waveguide and fabricated on a single-layer copper circuit board using a flame-retardant 4 substrate with a relative permittivity of 4.3 and a thickness of 1.6 mm. A prototype with compact dimensions of 0.040×0.040×0.0009 λ³ was constructed and experimentally evaluated. Measurements reveal a nearly omnidirectional radiation pattern, achieving peak gains of 2.92 dBi at 2.4 GHz and 4.25 dBi at 5 GHz. The antenna demonstrates a wide 10 dB return loss bandwidth of 67.7% (1.7–3.44 GHz) for the lower band and 56% (4.59–8.16 GHz) for the upper band. The strong agreement between simulated and measured results validates the design’s potential for practical and scalable implementation. This DBPA design offers a simpler, more compact, and wider-bandwidth alternative to conventional antennas, making it ideal for modern WLAN systems
Archimides multiband micro ribbon spiral antenna for energy harvesting
Full text linkThe current need to conserve natural resources and find new ways to advance without damaging the environment has led to the search for different energy sources. In this sense, the availability of radio frequency (RF) energy is found as a favorable option for new energy sources. This work describes the design of a microstrip antenna for the collection of radiofrequency energy in the megahertz band. Using the automatic optimization software computer simulation technology (CST) Studio, a circular spiral antenna is simulated using the low-cost FR4 substrate with a thickness of 1.57 mm and copper as the conductive material with a thickness of 0.035 mm. The proposed design presents resonance frequencies in multiple bands from 550 MHz to 1900 MHz, with bandwidths between 15 MHz and 150 MHz. The antenna design is based on the resonant cavity model and presents circular polarization due to its design coil type, with modified geometry using symmetrical orthogonal slots to generate multiple working bands. The design of this antenna can capture the power emitted by the frequency bands used in mobile telephony, radio communications, broadcasting, and television
Comparative performance analysis of convolutional neural network-architectures on coffee-bean roast classification
Full text linkThe classification of coffee bean roast levels using Agtron standards has evolved from traditional subjective methods to technology-driven approaches employing advanced artificial intelligence. Recent advancements in computer vision have demonstrated the capability of convolutional neural networks (CNNs) in providing objective and consistent roast level classification compared to human visual assessment, which is prone to variability and subjectivity. This research presents a performance analysis of five CNN architectures (AlexNet, ResNet, MobileNet, VGGNet, and DenseNet) for classifying coffee beans into eight distinct Agtron roast levels. The comprehensive methodology encompasses four phases: i) data acquisition, ii) image preprocessing, iii) model training and validation, and iv) evaluation metric. During training-validation, DenseNet outperformed other models, achieving 99.702% training accuracy and 77.68% validation accuracy. In the testing evaluation, DenseNet also led with an average testing accuracy of 93.8%, followed by ResNet at 92.6%, VGGNet and AlexNet both at 92.4%, and MobileNet at 89.7%. The results show that the DenseNet shows promise in classifying Agtron coffee-bean roast classification