Taiwan Association of Engineering and Technology Innovation: E-Journals
Not a member yet
887 research outputs found
Sort by
Development and Application of a Body Joint Angle Detection System for Free-Throw Shooting Prediction and Posture Correction
Get PDFThe success rate of free-throw shooting is often a critical factor in determining game outcomes. This study employs machine learning to develop a low-cost, hardware-free joint angle measurement system for free-throw shooting and applies it to the scientific training of free-throw shooting skills. With the system, the joint angle curves of players can be measured without the need for reflective markers, thereby reducing setup costs and facilitating the integration of scientific training. This study presents several innovative features. The experimental results indicate that the amount of training data required for modeling is 50% of that required by the J48 decision tree classifier, with an accuracy 1.2 times higher. Additionally, when a shot is missed, the system compares the disparity in joint angles and provides feedback for posture correction, allowing players to target specific problem areas for training, improve free-throw performance, and assist the team in winning games
Motorcycle Parking Violation Detection System Using YOLOv7 with Region of Interest Mapping and Object Area Calculation
Get PDFThe large number of motorcycle users has created challenges, particularly related to parking violations, which can lead to traffic congestion, hinder emergency access, disrupt pedestrian pathways, and inconvenience other users. Therefore, this study aims to detect motorcycle parking violations in unsupervised restricted areas using YOLOv7 to classify non-parking, parking, and personal objects. The best model is achieved at the 28th epoch with an mAP value of 0.953 at the 0.5 threshold. Parking restriction areas are defined using a Region of Interest (ROI), where violations depend on the parking object’s detected coverage within the ROI exceeding 50%. By employing an area calculation method, the results show better performance compared to methods without area calculation, achieving a recall of 89.7%, precision of 82.6%, and F1-score of 86.2% with a confidence threshold of 0.5
A Review of Advances in Bio-Inspired Visual Models Using Event-and Frame-Based Sensors
Get PDFThis paper reviews visual system models using event- and frame-based vision sensors. The event-based sensors mimic the retina by recording data only in response to changes in the visual field, thereby optimizing real-time processing and reducing redundancy. In contrast, frame-based sensors capture duplicate data, requiring more processing resources. This research develops a hybrid model that combines both sensor types to enhance efficiency and reduce latency. Through simulations and experiments, this approach addresses limitations in data integration and speed, offering improvements over existing methods. State-of-the-art systems are highlighted, particularly in sensor fusion and real-time processing, where dynamic vision sensor (DVS) technology demonstrates significant potential. The study also discusses current limitations, such as latency and integration challenges, and explores potential solutions that integrate biological and computer vision approaches to improve scene perception. These findings have important implications for vision systems, especially in robotics and autonomous applications that demand real-time processing
Investigation of Effects of Process Variables on Weld Bead Characteristics in Surface Coating of 309L Stainless Steel by Wire Arc Additive Manufacturing
Get PDFCoating carbon steel surfaces with stainless steel is a crucial technology in various industries to extend the product lifespan. This study focuses on investigating the effects of process parameters on weld bead characteristics in coating SS309L on carbon steel substrates by wire arc additive manufacturing (WAAM) and identifying the optimal parameters. The key parameters are current, travel speed, and voltage, while the weld bead characteristics include height, width, and depth of penetration. Experimental data and analysis of variance (ANOVA) are employed to develop and evaluate predictive models in Minitab software. The results show that the optimal process parameters for coating SS309L on carbon steel substrates by WAAM are voltage = 22 V, current = 132 A, and travel speed = 0.3 m/min, which improve height and width by 56.71% and 25.87%, respectively, while reducing the depth of penetration by 21.74% compared to the worst-case scenario
Formulating Seismic Intensity Scale (JMA-SIS) Using Response Spectrum: A New Approach for Structural Engineering Design
Get PDFThis study aims to formulate a calculation for earthquake shaking intensity (rs_mSIS) based on the response spectrum (RS) using the Japan Meteorological Agency-seismic intensity scale. The research investigates the relationship between the response spectrum parameters—period and maximum acceleration—and the earthquake source types, including megathrust, Benioff, and shallow crust/background sources. Artificial ground motions are generated and analyzed using Matlab to calculate shaking intensity values, which are then used to develop the rs_mSIS formula. The formulation is validated against actual response spectrum data from 15 Indonesian cities and demonstrated high accuracy, with the Wariyatno coefficient applicable across all models. This approach provides a standardized method to assess seismic intensity, offering enhanced reliability for building design in earthquake-prone areas and serving as a valuable tool for engineers and urban planners to improve earthquake resilience in diverse seismic environments
Assessing the Effectiveness of Exclusive Truck Lanes: A Korean Expressways Case Study
Get PDFThis study investigates the operational and safety impacts of introducing exclusive truck lanes on the Gyeongbu Expressway in South Korea, addressing the necessity of such lanes due to the disparities in vehicle weight and performance between trucks and passenger cars. The study focused on the 33.2 km, one-way, four-lane Chilgok-Mulryu to Gimcheon segment, adhering to international installation criteria. Using the micro-simulation tool VISSIM, the rightmost lane was modeled as exclusive for trucks, and traffic operations and safety were analyzed under varying conditions of Level of Service (LOS). Under LOS C, the exclusive truck lane reduced the speed standard deviation (a surrogate safety measure) with minimal travel speed reduction. Conversely, under LOS A with low traffic, average travel speed declined, and speed standard deviation increased due to limited truck usage. These findings highlight the need for flexible truck lane management tailored to traffic and road conditions
An Abductive Reasoning Approach for Energy Saving in Robotic Systems
Get PDFThe velocity and acceleration commands of industrial robots are set to their maximum values to shorten the cycle time of products. However, the excessively high speed and acceleration for movements can cause unnecessary mechanical energy and electricity consumption. This paper proposes an energy-saving approach for robotic systems based on abductive reasoning. Results for different combinations of speed commands and acceleration commands are evaluated based on energy consumption and cycle time. Moreover, a well-designed abduction rule formula is used to achieve a good balance between cycle time and mechanical energy consumption of industrial robots. Simulation results of a Franka robot by ROS, Gazebo, and Moveit verify the effectiveness of the proposed approach
Optimization Method for Cross-Regional Scheduling of Retired Charging Pile Component Reuse
Get PDFWith the increasing number of decommissioned charging piles, efficient reuse of their components is essential for sustainable resource utilization and intelligent grid management. To address the challenges in recycling and scheduling retired charging pile components, this study proposes a cost-optimization approach for delivery planning in smart grid logistics. An Electric Vehicle Routing Problem with Time Windows (EDVRP-TW) model is formulated that considers vehicle capacity and time constraints. To solve it, an Improved Chicken Swarm Optimization Algorithm (ICOOT) is developed, integrating Circle chaotic mapping, spiral search strategy, and normal cloud mutation to enhance convergence speed and solution quality. Simulation experiments using real-world datasets demonstrate that the proposed method significantly reduces operation and maintenance costs, achieving up to an 11.77% cost reduction. The results validate the effectiveness and applicability of the model and algorithm in intelligent recycling and scheduling of grid materials
Study of Industrial Accident Based on In-Depth Investigation
Get PDFIndustrial accidents caused by static electricity were common for years. Beyond grounding and other static mitigation devices, plants often control environmental humidity and optimize production processes to reduce static hazards. This study aims to determine the causes and mechanism of an industrial accident and well-known unusual event related to static electricity, volatile organic compounds (VOC), and minimum ignition energy (MIE). Through instrument measurements and analysis methodology such as Hartmann tube, static electricity meter and fault tree, the cross factors are analyzed to complete the study systematically. Results reveal that anti-rust paint on the inner surfaces of machinery created insulating conditions, allowing static electricity to accumulate and discharge during material feeding. The induced static electricity subsequently releases a high discharge energy exceeding the MIE of the particle. When combined with VOC generation from an unexpected process interruption, it can lead to fire or explosion
Hardware Implementation of Chaotic Image Encryption on FPGA
Get PDFThis paper proposes a new FPGA-based architecture for secure image encryption, utilizing chaotic maps for permutation and substitution mechanisms. Specifically, Duffing map is employed to generate permutation addresses, and the Henon map for value substitution; both techniques produce pseudo-random sequences with strong sensitivity to initial conditions. The hardware design implemented using Xilinx System Generator and deployed on an Artix-7 FPGA supports both grayscale and RGB image formats. A comprehensive performance analysis, utilizing Mean Squared Error, Histogram analysis, Correlation Coefficient Analysis, Number of Pixels Change Rate, and Unified Average Changing Intensity, demonstrates perfect image recovery and robust encryption resistance. The keys used in the encryption system are passed using the NIST STS randomness test. The single-round and multi-round designs deliver a processing acceleration ranging from 2.11 to 3.92 for image sizes 128×128 to 1024×1024 pixels, highlighting their effectiveness and practicality for real-time encryption in low-latency environments