Publication Search

The scarcity of real-world data in Air-Conditioning (AC) fault diagnosis necessitates the use of synthetic data; however, rule-based synthetic datasets often suffer from a significant sim-to-real domain gap. To address this, we propose a Model-Data Coevolution (MDC) framework that employs a Simulated Annealing (SA) controller to optimize augmentation parameters. We introduce a novel technique, Stochastic Feature Decoupling (SFD), which applies independent noise to raw and derived features, contrasting it with traditional Logically-Consistent Augmentation (LCA). Empirical results show that SFD significantly outperforms LCA, achieving a weighted F1-score of 0.93 and increasing NORMAL class recall to 82%. We demonstrate that by breaking deterministic feature links, SFD acts as a robust regularizer, utilizing "physically impossible" data to enhance generalization in complex real-world environments.

Plastic waste detection in indoor environments is an essential challenge in the development of intelligent cleaning systems and robotic automation. Small and medium-sized plastic debris is often difficult to identify using conventional methods due to variations in color, shape, and reflectance. This study proposes an image-processing-based approach that combines thresholding and contour detection techniques to improve the accuracy of detecting plastic objects on floor surfaces. The initial stage involves converting the image into a color space that is more stable under varying illumination, such as HSV or grayscale, to reduce the influence of lighting intensity. Subsequently, adaptive thresholding is applied to separate plastic objects from the background by using dynamic threshold values tailored to the image’s conditions. The segmentation results are refined through morphological operations such as opening and closing, enabling the removal of small noise and enhancing the clarity of object boundaries. The core stage of the system employs contour detection to extract object shapes and areas, allowing the identification of plastic waste based on size, perimeter, and specific geometric characteristics. Experiments were conducted under different lighting conditions and various floor types, and the results demonstrate that the proposed approach successfully detects plastic debris with satisfactory accuracy and relatively fast processing time. Therefore, this method is suitable for implementation in robotic cleaning systems, indoor cleanliness monitoring devices, and other computer vision applications requiring real-time and efficient object detection.

Accurate wind turbine power curve modeling plays a crucial role in performance evaluation, energy yield estimation, and data-driven control strategies. However, actual power curves often exhibit non-linear behavior influenced by atmospheric variability, measurement noise, and SCADA anomalies, making conventional modeling approaches less effective. This study proposes an optimized logistic power curve model whose parameters are tuned using Particle Swarm Optimization (PSO) to improve predictive accuracy. The analysis uses the Wind Turbine SCADA Dataset from Kaggle, which undergoes extensive preprocessing including physical rule filtering, outlier detection with the Interquartile Range (IQR) method, anomaly removal, and smoothing of the power signal. A three-parameter logistic model is selected due to its ability to capture the typical S-shaped relationship between wind speed and power output. PSO is applied to identify optimal model parameters by minimizing the Mean Squared Error (MSE), utilizing 40 particles over 200 iterations. The optimized model achieves strong predictive performance with RMSE of 404.09, MAE of 179.96, and R² of 0.904 on the test set, indicating that more than 90% of the variability in actual power can be explained by wind speed. Residual analysis reveals heteroscedastic patterns and slight overestimation in mid-range wind speeds, yet overall model consistency remains high. Comparative evaluation against Linear Regression, Random Forest, and logistic modeling using curve_fit shows that the Logistic–PSO approach provides the most accurate and stable predictions. These findings demonstrate that combining logistic modeling with PSO offers an effective and robust method for data-driven wind turbine power curve optimization.

An exhaust pipe is a tubular device used to channel combustion gases from a vehicle’s engine into the environment. In addition to this primary function, the exhaust also serves to reduce the noise level produced by engine combustion. The component of the exhaust system that significantly affects torque, brake mean effective pressure (BMEP), and noise level is the silencer. This study aims to compare the torque, BMEP, and noise levels produced by variations in the length and construction dimensions of elliptical filter designs in the silencer. The silencers used in this research have lengths of 280 mm, 300 mm, and 260 mm, with short ellipse filter diameters of 25 mm and long ellipse filter diameters of 30 mm, 40 mm, and 50 mm. The research employs an experimental quantitative method, and the data were analyzed using one-way Analysis of Variance (ANOVA). The experiment was conducted in a Mechanical Engineering workshop using a Yamaha R15 V3 155cc injection motorcycle (2021), a Super Dyno 50L dyno test, and a sound level meter, from February to April 2024. The engine speeds tested were 1500, 2500, 3500, 4500, 5500, and 6500 rpm. The results show that the variation of ellipse dimensions 25 mm × 50 mm × 260 mm produced a maximum torque of 12.77 N·m at 4500 rpm, a maximum BMEP of 1021 kPa at 4500 rpm, and a noise level of 80.3 dB. The variation 25 mm × 40 mm × 300 mm produced a maximum torque of 12.88 N·m, a BMEP of 1042 kPa, and a noise level of 75.60 dB, while the variation 25 mm × 30 mm × 280 mm produced a maximum torque of 12.67 N·m, a BMEP of 1013 kPa, and a noise level of 75.63 dB.

This study ais to formulate development strategies for Yoga Barbershop in Palaran, East Kalimantan, using SWOT analysis and the Quantitative Strategic Planning Matrix (QSPM). SWOT analysis is utilized to assess internal and external factors,combned with QSPM to determine priority strategies. The study reveals that Yoga Barbershop’s strengths nclude customer loyalty, a strategic location, modern haircut styles,and ownership of the business premises, which reduces operational costs.However, identified weaknesses consist of limited service facilities, inflexible operating hours, restricted space, noise issues, and insufficient staff. Meanwhile, opportunities for Yoga Barbershop include opening new braches and recruiting additional employees. Threats arise from similiar business competitors and changes in hairstyle trends. Based on the findings, Yoga Barbershop’s IE matrix places the business in quadrant V, with a total IFE score of 2,72 and an EFE score of 2,71, indicating a stable position that requires stability strategies. The recommended strategy is the weakness-opportunity (WO) approach. Priority strategic recommendations, in order of importance, are opening a new branch, training new staff, and increasing the flexibility of operational hours.

Wireless communication, in its infrastructure nature, faces many challenges such as fading, data coverage, and interference issues. Therefore, High-Fidelity or (Li-Fi) is utilized due to its ability to naturally provide high-density wireless data coverage in closure’s particularly helpful for application(s) in some areas while the radio interference conditions are concern. This article illustrates an advanced Li-Fi approach performing high-speed data transmission between two Personal Computers (PCs) utilizing the Arduino Nano-based technique. In the experimental phase, data is mainly used to be transmitted over red laser diode (630 nm) through (30 cm) in distance, a distance of 30 cm, achieving a high peak speed reach to about (512Bps). The proposed approach performance is computed by evaluating the most important and related metrics like Signal-to-Noise Ratio (SNR), Bit-Error-Rate (BER), and influence of throughput on input data over various light circumstance. The proposed approach mainly utilizes a keypad as a user input and two related detection models for both a solar cell and a photodetector in order to make a powerful comparison in terms of performance. the results showed that when the photodetector applies a higher-detection efficiency (via BER enhancement which reaches to 20% over solar-cell), the solar-cell clarify outstanding power and cost-activity. The mentioned findings are propped by elaborated statistical-analyses and MATLAB simulation to design, simulate and visualize the validate functionalities of the robustness and scalability properties of the proposed Li-Fi approach.

Quantum sensing technologies are revolutionizing precision measurements in electrical engineering by leveraging quantum phenomena to achieve unprecedented accuracy. This paper examines the applications of quantum sensors in electric field detection, biomedical instrumentation, and power grid monitoring. The study explores their ability to enhance sensitivity, minimize noise interference, and improve real-time data acquisition. Through advanced quantum principles, these sensors enable high-resolution diagnostics in medical devices, optimize energy efficiency in smart grids, and enhance electromagnetic field measurements. The findings highlight the transformative potential of quantum sensors in modern engineering applications, paving the way for smarter and more efficient monitoring systems.

Optical Character Recognition (OCR) systems often struggle to extract text accurately from images captured at various distances, particularly under challenging conditions such as blurriness, noise, or poor lighting. These issues are common in real-world scenarios and limit the effectiveness of existing OCR technologies. This study addresses these challenges by applying Gaussian blur after the grayscale conversion. This method reduces noise for the image's clarity without sacrificing the original algorithm's key features. Results revealed that the enhanced OCR algorithm significantly outperformed existing methods in terms of accuracy and confidence levels. It demonstrated the ability to read signages with higher precision, even in difficult conditions such as intricate designs, poor lighting, and long distances. This advancement enables more reliable text recognition and translation, offering practical applications for public signage translation, cross-cultural communication, and improved accessibility in multilingual environments.

Signature verification is crucial for confirming the authenticity of identities in both administrative and financial transactions, where signature forgery can lead to significant security risks. The Harris Corner Detector Algorithm is a widely used method for feature extraction in image processing; its application spans various domains, such as detection of signature forgery. While effective in identifying key features, noise significantly affects performance, especially with impulse noise like salt-and-pepper noise commonly found in signature images. To solve this problem, this study enhances the Harris Corner Detector Algorithm by applying a median filter before gradient calculation. This method removes noise without sacrificing the integrity of key features important in signature forgery detection. The study evaluates the original and the enhanced algorithm using standard image quality metrics. Peak Signal-to-Noise Ratio (PSNR) surged from an average of 13.6 dB to 43.28 dB, the Structural Similarity Index (SSIM) improved significantly from 78% to 94%, and the Mean Squared Error (MSE) dropped substantially from 16.74 to 3.84. These advancements resulted in a more reliable algorithm, exhibiting excellent resistance to noise while maintaining image structure, making the enhanced algorithm highly effective for accurate signature forgery detection.

Noise in libraries can disrupt the concentration and comfort of visitors who are reading or studying. Therefore, a tool is needed to detect noise and automatically issue warnings to maintain silence in the library environment. This research aims to design and develop a noise detection tool in libraries using Internet of Things (IoT) technology. The device consists of a sound sensor that detects noise levels and sends data to a cloud-based server for analysis. When the noise level exceeds a specified threshold, the tool provides warnings through visual indicators and notifications to the library staff. The prototype was tested in a library environment and was able to detect noise accurately and issue real-time alerts. The results showed that the device effectively reduced noise levels and improved the comfort of library visitors. With the implementation of IoT technology, this system can be accessed and monitored remotely, facilitating library management in maintaining a conducive environment.

This research produces a monitoring system to determine the position of the ship using the Kalman Filter method in web-based system development using the waterfall development method. This system allows users to view the position and coordinates of the ship in real-time. By applying the Kalman Filter method, uncertainty and noise in measuring the position of the ship can be reduced, thereby increasing the accuracy in determining the actual position of the ship. The system development process is carried out using the waterfall method which consists of requirements analysis, design, implementation, testing, and maintenance. The developed web-based system provides more accurate and reliable information to users, with the ability to view the position and coordinates of the ship in real time. This research contributes to the development of a more effective ship position monitoring system that can be used in various purposes such as navigation, surveillance, and ship monitoring.

An important digital image processing is image segmentation, which separates objects from the background for further analysis. One segmentation technique is edge detection, which looks for boundaries between areas of different brightness. This article compares four edge detection methods: Roberts, Prewitt, Sobel, and Canny. The results show that, despite requiring more complex computations, Canny's method produces the sharpest and best connected edges; Sobel and Prewitt's method, on the other hand, is faster and simpler than Roberts' method, but is less effective in dealing with noise and often produces edges that are not connected to the plane. The choice of edge detection method depends on the application. Sobel and Prewitt are good for speed and stability, and Roberts is suitable for fast processing of images with minimal noise.

implemented and evaluated the edge detection method using the Sobel Operator, which calculates the gradient of image intensity through two convoluted kernels for horizontal (Gx) and vertical directions. (Gy). The magnitudo gradient is obtained from the combination of both such directional gradients to represent the edge force on each pixel. The main steps include image pre-processing, the application of the Sobel kernel, the calculation of magnitudo gradients, and the filtering of results to extract significant edges. The results show that the Sobel Operator is effective in highlighting intensity differences that indicate the boundary of the object, although it is sensitive to noise and less accurate for fine edges. Despite its limitations, this method is simple to implement and useful as an initial step in image processing applications such as segmentation, pattern identification, and object shape analysis.