Taiwan Association of Engineering and Technology Innovation: E-Journals
Not a member yet
887 research outputs found
Sort by
Enhanced Vehicle Dynamics through Constrained Model Predictive Control for In-Wheel Active Suspension Systems
Full text linkThis study develops and evaluates a Model Predictive Control (MPC) strategy to enhance the dynamic performance of vehicle suspension systems subjected to stochastic road excitation. A high-fidelity simulation environment is established using a quarter-car model and a road profile conforming to ISO 8608:2016 Class B standards, with a constant vehicle speed of 60 km/h. The proposed constrained MPC algorithm, which predicts system states and optimizes control inputs over a finite horizon, is benchmarked against a conventional passive suspension. Simulation results demonstrate the MPC controller's superior efficacy in attenuating vibrations across a broad frequency range, resulting in significant improvements in both ride quality and handling stability. Key performance metrics include a 36.44% reduction in sprung mass acceleration (enhancing passenger comfort), an 18.71% reduction in unsprung mass acceleration (improving vibration isolation), and a 6.02% reduction in hub acceleration (promoting stable tire-road contact)
Real-Time Video-Based Posture Monitoring Measurement of Back Angles Using YOLOv8 and Edge Detection for Strength Training
Full text linkWorkout-related lower back injuries are common during strength training and are often caused by improper posture, highlighting the need for real-time posture monitoring to support injury prevention and performance optimization. This study proposes a mobile-based computer vision approach for real-time quantification of back angles during workouts. The proposed method integrates YOLOv8 instance segmentation to isolate the trunk region and applies Canny edge detection for contour extraction. It then employs a geometric formulation to identify neck and back reference points for angle computation. This hybrid design enables robust trunk localization and stable angle estimation across dynamic exercise movements. The model is evaluated on six gym-recorded videos captured with a low-cost mobile camera, achieving a mean relative error of 6.49%, comparable to video-based biomechanical assessment methods. These findings indicate that the proposed approach provides an efficient and practical solution for real-time back-posture monitoring using mobile devices, supporting safer and higher-quality daily training
AI-Driven Anomaly Detection in Quadcopters Using ADXL345 Accelerometer Vibration Data and IoT Integration
Full text linkThis study investigates artificial intelligence methods for offline anomaly detection in quadcopters to improve flight safety. Vibration data were collected using ADXL345 accelerometers interfaced with ESP32 modules. Eight time-domain features were extracted from triaxial acceleration signals. Four machine learning classifiers—Random Forest (RF), Support Vector Machine, K-Nearest Neighbors, and Neural Networks—were trained and evaluated on a dataset representing a healthy state and four propeller damage levels (10% to 40% cuts). The RF classifier achieved the highest accuracy of 98% using standard deviation features. The results demonstrate the effectiveness of time-domain features and tree-based models for propeller fault diagnosis. This benchmarking approach enables precise identification and quantification of propeller damage severity, supporting rapid maintenance decisions and proactive flight risk management for UAV platforms
Enhancing Visual SLAM Robustness in Dynamic Scenes with YOLOv5-Assisted ORB-SLAM3
Full text linkThis study presents an enhanced visual SLAM (Simultaneous Localization and Mapping) framework that integrates ORB-SLAM3 with the YOLOv5 real-time object detection model to improve pose accuracy in dynamic environments. Although ORB-SLAM3 achieves robust performance in static scenes, its reliance on ORB feature tracking often degrades accuracy in the presence of moving objects. To overcome this limitation, YOLOv5 is employed to identify dynamic regions in each video frame, enabling the system to remove motion-related feature points before matching. This filtering mechanism reduces the influence of dynamic objects on trajectory estimation and enhances overall system robustness. The proposed method was evaluated using dynamic datasets, including BONN and TUM RGB-D, and further validated through real-world experiments with an Intel RealSense D435i camera. Experimental results demonstrate substantial improvements in pose accuracy compared with the baseline ORB-SLAM3 and the RTAB-Map system, confirming the effectiveness of the YOLOv5-assisted ORB-SLAM3 integration in dynamic scenes
Strength Prediction of Rectangular FRP-Reinforced Concrete Columns Under Eccentric Loading
Full text linkThis study aims to develop an analytical model for evaluating the load-carrying capacity of rectangular fiber-reinforced polymer (FRP) reinforced concrete columns under eccentric loading. In the proposed model, the contribution of FRP bars in compression is considered, with their compressive strength estimated as a fraction of tensile strength. Meanwhile, the effects of confinement, tension stiffening, and second-order effects are conservatively neglected. Two main failure modes, namely concrete crushing and FRP rupture, are distinguished by the balanced failure condition. This model applies strain compatibility with the plane section and constitutive laws to derive stress-strain distributions across the cross-section. Then, the model is validated against 91 experimental results covering diverse sections, strengths, and eccentricities (e/h = 0.1-1.0), showing high accuracy (mean: 0.932; RMSE: 0.154; COV: 22.9%; SD: 0.145; r: 0.84) and outperforming ACI CODE-440.11. Analysis results also show that compressive FRP reinforcement contributed between 0.94% and 22.3% to the column strength
Material Development and Properties of Medium-Density Board from Low and High-Density Polyethylene
Full text linkThis study aims to develop a material from waste low-density polyethylene (LDPE) and high-density polyethylene (HDPE) into a medium-density board and assess its mechanical and physical properties. The development starts with degreasing the upcycled plastic sheets, stacking using premixed polyester resin as an adhesive, pressing, and laminating. The specimens are sent to the Department of Science and Technology Industrial Technology Development Institute (DOST-ITDI) standards and testing division to determine the material’s mechanical and physical properties. The findings reveal that the medium-density board successfully combines LDPE and HDPE waste, achieving tensile, flexural, and compressive strengths of 12.1 MPa, 24.2 MPa, and 14.5 MPa, respectively. The board is suitable for shaded outdoor use but not for continuous immersion as it shows a heat deflection temperature of 57.8 ℃ and 1.27% water absorption after 24 hours. Therefore, it is a potential substitute for furniture, home decor, and light construction materials
Coordinated FCS-MPC and Auxiliary Damping Control for Enhanced SSR Mitigation in Series-Compensated DFIG Wind Farms
Full text linkThis study proposes a comprehensive and novel coordinated control framework to effectively mitigate sub-synchronous resonance (SSR) in a doubly-fed induction generator based on wind turbines connected to series-compensated transmission lines. The proposed approach integrates finite control model predictive control in the rotor-side converter to achieve fast and accurate current tracking, while an SSR damping controller is embedded in the grid-side converter (GSC). Through coordinated operation, the GSC injects an auxiliary damping signal into the q-axis current reference to suppress sub-synchronous oscillations while maintaining DC-link voltage stability. Time-domain simulations conducted under diverse operating conditions demonstrate the robustness and superior performance of the proposed scheme. In particular, the total harmonic distortion is reduced from over 2.3% to below 2.0%, and the maximum electromagnetic torque oscillation is significantly suppressed from over 1.0 pu to approximately 0.02 pu, thereby confirming the effectiveness of the proposed control strategy in enhancing overall system stability
Modeling and Analysis of a Tensegrity-Based Vibratory Platform Driven by Piezoelectric Actuators Using IronCAD
Full text linkThis study aims to perform a simulation study of a tensegrity-based vibratory platform driven by piezoelectrical actuators using IronCAD software. The platform is capable of advancing parts in any specified direction or rotation on the horizontal plane. The platform’s structure is presented first. Then, the proposed platform's solid model is established using IronCAD software. Moreover, piezoelectric actuators are modeled in Multiphysics for IronCAD by specifying the piezoelectric material properties. Various inputs to the platform are simulated and investigated. The simulation results demonstrate the effectiveness of IronCAD for modeling and analyzing the proposed design
Improved CNN-LSTM Bearing Remaining Useful Life Prediction Based on the Weibull Loss Function
Full text linkThe prediction of the remaining useful life (RUL) of rolling bearings is a critical task in predictive maintenance. This paper presents a deep learning model named knowledge-driven convolutional neural network–long short-term memory (KCNN-LSTM), enhanced by the Weibull-based loss function tailored with historical bearing failure data. By incorporating a probabilistic Weibull modeling mechanism, the model can accurately capture the uncertainty and accelerated degradation trend of bearing failure over time. The prognostics and health management (PHM) 2012 and XJTU-SY bearing datasets are utilized to evaluate the proposed KCNN-LSTM model. The results indicate that the proposed KCNN-LSTM achieves superior performance compared with the conventional CNN-LSTM model, leading to a 10.2% improvement in the score metric and a notable reduction in prediction error. The proposed model offers a practical and effective approach for enhancing predictive maintenance in high-reliability industrial systems
Assessment of Electromagnetic Exposure Level in an Industrial Facility: EM Pollution Level of Employees
No full textWith the rapid advancement of technology, residential and occupational electromagnetic field (EMF) exposure has become an important issue. The electromagnetic exposure level in residents and working areas should be monitored and controlled to provide a healthy environment. This study focuses on assessing the EMF exposure levels in a factory with 250 employees. Electric and magnetic fields measurements are conducted from offices to manufacturing areas, specifically within the extremely low-frequency and radio frequency bands. Then, the results are compared with the reference levels set by the International Commission on Non-Ionizing Radiation Protection (ICNIRP) and the national Information and Communication Technologies Authority of Türkiye (ICTA: BTK in Türkiye). The findings indicate that EMF levels throughout the working environment are below the ICNIRP and ICTA reference levels. Consequently, the facility is classified as safe regarding electromagnetic exposure under the observed conditions