Jurnal Nasional Teknik Elektro
Not a member yet
359 research outputs found
Sort by
Speed Control of an Electrical Cable Extrusion Process Using Artificial Intelligence-Based Technique
Full text linkMost cable manufacturing companies use Programmable Logic Controllers with conventional controllers to control line speed during cable extrusion. These traditional controllers have difficulties keeping the line speed constant, causing surface defects on the extruded cables and affecting the quality of the manufactured cables. To overcome these challenges, data on the causes of defects during cable manufacturing were collected from a cable manufacturing company in Ghana to ascertain the possible causes during cable manufacturing. Adaptive Neuro-Fuzzy Inference System (ANFIS) controller was designed to provide a constant line speed during the cable extrusion process. To ascertain its robustness, the ANFIS controller was compared to a conventional Proportional Integral Derivative controller and a Fuzzy Logic controller. The controllers were designed and simulated using MATLAB/Simulink software. The analysis of the collected data indicated that a break in insulation/ sheath was a frequently occurring defect during the cable manufacturing process due to improper line speed control of the machines used in the cable manufacturing process. Based on the results obtained from the various controllers, it was concluded that the ANFIS controller was robust in achieving stability regarding line speed variations
Enhancing Interface Efficiency: Adaptive Virtual Keyboard Minimizing Keystrokes in Electrooculography-Based Control
Full text linkRapid technological developments, one of which is technology to build communication relationships between humans and machines using Biosignals. One of them is Electrooculography (EOG). EOG is a type of biosignals obtained from eye movement. Research related to EOG has also developed a lot, especially for virtual keyboard control. Research on virtual keyboard control based on eye gaze motion using electrooculography technology has been widely developed. Previous research mostly drew conclusions based on time consumption in typing paragraphs. However, it has not been seen based on the number of eye gaze motions made by the user. In this research, an adaptive virtual keyboard system is built, controlled using EOG signals. The adaptive virtual keyboard is designed with 7x7 dimensions and has 49 buttons, including main buttons, letters, numbers, symbols, and unused buttons. The layout of the adaptive virtual keyboard has six zones. Each zone has a different number of steps. Characters located in the same zone have the same number of steps. The adaptive feature is to rearrange the position of the character's button based on the previously used characters. In the experiments, 30 respondents controlled static and adaptive virtual keyboards with 7 paragraphs typed. Adaptive mode rearranges the position of buttons based on k-selection activities from respondents. the k numbers are 10, 30, 50, 70 and 100. Two virtual keyboard modes are evaluated based on the number of steps required to type the paragraphs. Test results show that the performance of the adaptive virtual keyboard can shorten the number of user steps compared to static mode. There are tests of the optimal system that can be reduced up to 283 number of steps and from respondents, that can reduced up to 258 number of steps or about 40% of steps. This research underscores the promise of EOG-driven adaptive virtual keyboards, signaling a notable stride in augmenting user interaction efficiency in typing experiences, heralding a promising direction for future human-machine interface advancements
Fault Detection and Diagnosis of a 3-Phase Induction Motor Using Kohonen Self-Organising Map
Full text linkThis paper uses the Kohonen Self-Organising Map (KSOM) to detect, diagnose, and classify induction motor faults. A series of simulations using models of the 3-phase induction motor based on real industrial motor parameters were performed using MATLAB/Simulink under fault conditions such as inter-turn, power frequency variation, over-voltage and unbalance in supply voltage. The model was trained using the input signals of the various fault conditions. Various faults from an unseen induction motor were fed to the model to test the model’s ability to detect and classify induction motor faults. The KSOM adapted to the conditions of the unseen motor, detected, diagnosed and classified these faults with an accuracy of 94.12%
Hyperbolic Tangent - Based Adaptive Inertia Weight Particle Swarm Optimization
Full text linkThis paper presents a study on using adaptive inertia weight (AIW) in particle swarm optimization (PSO) for solving optimization problems. An AIW function based on the hyperbolic tangent function was proposed, with the function parameters adaptively tuned based on the particle best and global best values. The performance of the proposed AIW-PSO was compared with standard PSO and other PSO variations using seven benchmark functions. The results showed that the proposed AIW-PSO outperformed the other variations in terms of minimum cost and mean cost while reducing the standard deviation of cost. The performance of the different PSO variations was also analysed by plotting the best cost against iteration, with the proposed AIW-PSO showing a faster convergence rate. Overall, the study demonstrates the effectiveness of using an adaptive inertia weight function in PSO for optimizing problems
Performance Process of Coil Winding Machine Based on Accuracy and Speed for Water Pump Motor
Full text linkA coil winding machine for water pumps using a monitoring system is a development of conventional winding tools. In regular coil winding tools, the coil winding process is done manually by rotating the handle as many times as the desired number of turns. The conventional winding tools have problems consisting of inconsistent working speed and operator-dependent winding continuity. Undesirable windings can occur with conventional winding tools, and the winding process requires close supervision. Therefore, the automatic coil winding machine was developed to optimize the coil winding process. The machine utilizes a DC motor to rotate the coil rolls, replacing the conventional roller handle function. This machining method uses an optocoupler sensor. The sensor serves to identify and evaluate the rotation of the roller. In addition, the ATmega8 microcontroller was applied to develop a system that can work automatically. Data collection involves varying the number of wire turns and the wire diameter dimension. The variation is necessary because the number of windings and wire diameter affect pump efficiency and performance. The data testing showed a machine accuracy rate of 98%, with a maximum difference of 1 coil winding in the results. This data confirms that the coil winding machine meets the tool's accuracy standards
Innovative Personal Assistance: Speech Recognition and NLP-Driven Robot Prototype
No full textThis paper presents the development and evaluation of a personal assistant robot prototype with advanced speech recognition and natural language processing (NLP) capabilities. Powered by a Raspberry Pi microprocessor, it is the core component of the robot's hardware. It is designed to receive commands and promptly respond by performing the requested actions, utilizing integrated speech recognition and NLP technologies. The prototype aims to enhance meeting efficiency and productivity through audio-to-text conversion and high-quality image capture. Results show excellent performance, with accuracy rates of 100% in Indonesian and 99% in English. The efficient processing speed, averaging 9.07 seconds per minute in Indonesian and 15.3 seconds per minute in English, further enhances the robot's functionality. Additionally, integrating a high-resolution webcam enables high-quality image capture at 1280 x 720 pixels. Real-time integration with Google Drive ensures secure storage and seamless data management. The findings highlight the prototype's effectiveness in facilitating smooth interactions and effective communication, leveraging NLP for intelligent language understanding. Integrating NLP-based speech recognition, visual documentation, and data transfer provides a comprehensive platform for managing audio, text, and image data. The personal assistant robot prototype presented in this research represents a significant advancement in human-robot interaction, particularly in meeting and collaborative work settings. Further refinements in NLP can enhance efficiency and foster seamless human-robot interaction experiences
Electromedical Device And Expert System for Early Detection of Hyperemesis Gravidarum
Full text linkHyperemesis Gravidarum (HG) is a pregnancy complication that is often overlooked as it is typically considered normal. If HG is not properly treated, nutrition will not be fulfilled which can negatively affect maternal and fetal health and even maternal and fetal death. The exact cause of HG is not identified, so there are no effective preventive methods. However early detection can help for prompt and appropriate treatment. Therefore, a monitoring system for pregnancy conditions was designed for HG early detection. This system employs the MPX5050 DP pressure sensor for measuring blood pressure, the MAX30100 for assessing maternal heart rate and oxygen saturation, the MAX4466 sensor for monitoring fetal heart rate, and an expert system using the certainty factor method to diagnose the probability of hyperemesis gravidarum. The expert system achieves an accuracy of 93.33%. In comparison to the aneroid sphygmomanometer, the designed sphygmomanometer reveals a mean difference of 3.5 mmHg for diastolic pressure, with a standard deviation below 8 mmHg for both systolic and diastolic pressures. The measurement of heart rate and oxygen saturation has a deviation of 1.8 % and 1.02 % respectively. These deviations align with the standards specified by the Ministry of Health for medical devices. For the fetal heart rate, the mean deviation is 3.4 bpm, and the measurement error is 2.38%. Thus, this system can be utilized to monitor pregnancy conditions, enabling the early detection of hyperemesis gravidaru
External Leakage Current Separation to Determine Arrester Condition Due to Contamination
Full text linkLeakage current measurements can be used to determine the aging condition of the ZnO arrester. The leakage current that occurs in the arrester is divided into two, namely external and internal leakage currents. The external leakage current is affected by contamination and the internal leakage current is affected by the aging of the varistor in the arrester. The external and internal leakage currents are measured separately to determine their contribution to the arrester condition. In this study, the effect of salt contamination on the arrester was studied further. The level of contamination used consisted of low, medium and heavy. The obtained leakage current is analyzed using wavelet energy. The results of this study indicate that the wavelet energy of each leakage current is different and can be used as an indicator in further analysis. The conclusion obtained is that the external leakage current is affected by contamination and has a different energy with the internal leakage current due to aging of the varistor arrester components
ANN-Based Electricity Theft Classification Technique for Limited Data Distribution Systems
Full text linkElectricity theft has been a challenge for distribution systems over the years. Theft presents a massive cost to the system operators and other issues such as transformer overloading, line loading, etc. It has become crucial for measures to be implemented to combat illegal electricity consumption. This work sought to develop an artificial neural network-based electricity theft classifier for distribution systems with limited data, i.e., systems that can only provide consumption data alone and no auxiliary data. First, a novel data pre-processing method was proposed for the systems with consumption data only. Again, synthetic minority oversampling is employed to deal with the unbalance problem in the theft detection dataset. Afterwards, an artificial neural network (ANN)-based classifier was proposed to classify customers as normal or fraudulent. The proposed method was tested on actual electricity theft data from the Electricity Company of Ghana (ECG) and its performance compared to random forest (RF) and logistic regression (LR) classifiers. The proposed ANN-based classifier performed exceptionally by producing the best results over RF and LR regarding precision, recall, F1-score, and accuracy of 99.49%, 100%, 99.75%, and 99.74%, respectively.
 
Short-Term EV Charging Demand Forecast with Feedforward Artificial Neural Network
Full text linkThe global increase in greenhouse gas emissions from automobiles has brought about the manufacture and usage of large quantities of electric vehicles (EVs). However, to ensure proper integration of EVs into the grid, there is a need to forecast the charging demand of EVs accurately. This paper presents a short-term electric vehicle charging demand forecast using a feedforward artificial neural network optimized with a modified local leader phase spider monkey optimization (MLLP-SMO) algorithm, a proposed variant of spider monkey optimization. A proportionate fitness selection is employed to improve the update process of the local leader phase of the spider monkey optimization. The proposed algorithm trains a feedforward neural network to forecast electric vehicle charging demand. The effectiveness of the proposed forecasting model was tested and validated with electric vehicle public charging data from the United Kingdom Power Networks Low Carbon London Project. The model's performance was compared to a feedforward neural network trained with particle swarm optimization, genetic algorithm, classical spider monkey optimization, and two conventional forecasting models, multi-linear regression and Monte Carlo simulation. The performance of the proposed forecasting model was assessed using the mean absolute percentage error of forecast and forecasting accuracy. The model produced a forecast accuracy and mean absolute percentage error of 99.88% and 3.384%, respectively. The results show that MLLP-SMO as a trainer predicted better than the other forecasting models and met industry standard forecast accuracy.