Article
Computer Science, Hardware & Architecture
Hao Wu
Summary: This study focuses on classifying Indian classical dance using FPGA, with data collection on various themes and postures for offline reference in dance education.
MICROPROCESSORS AND MICROSYSTEMS
(2021)
Article
Computer Science, Hardware & Architecture
Jin Liu
Summary: The proposed method presents a reliable approach for estimating the three-dimensional pose of the human head in a single image by using data on skin and hair regions. It utilizes a Neural Network and efficient algorithm and offers improved robustness compared to techniques based on tracking facial features.
MICROPROCESSORS AND MICROSYSTEMS
(2021)
Article
Computer Science, Hardware & Architecture
Kai Zhang
Summary: This study investigated the impact of Animation visualization methods in FPGA pre-service kindergarten Embedded System on teaching Animation, and found that the use of Animation visualization methods can significantly increase participants' self-efficacy in teaching Animation.
MICROPROCESSORS AND MICROSYSTEMS
(2021)
Article
Computer Science, Information Systems
Jinda Xu, Meili Lu, Zhen Zhang, Xile Wei
Summary: This paper proposes a novel bio-inspired self-repairing CPG model based on tripartite synapse, and demonstrates its robustness and self-repairing capabilities through experiments in the presence of synaptic failures.
Article
Engineering, Electrical & Electronic
Carlos E. C. Souza, Davi Moreno, Daniel P. B. Chaves, Cecilio Pimentel
Summary: This work proposes a method to generate 1-D pseudo-chaotic sequences based on the discrete Arnold's map defined over the integer ring Z(2)m. The period of the generated sequences is investigated using properties of the Fibonacci sequence over Z(2)m. The statistical properties of the PRNG are analyzed, and it is shown that the proposed PRNG has higher throughput and competitive hardware consumption compared to other architectures in the literature.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Article
Computer Science, Hardware & Architecture
Mingquan Luo
Summary: Human Resource Management (HRM) is a structured process involving setting up organizations, managing finances and benefits, performance evaluation, and talent retention. Research programs are strategic tools for achieving core financial goals by evaluating employee performance through a performance management system. The implementation of a Bayesian Network Representation based Field Programmable Gate Arrays (FPGA) is proposed for continuous monitoring of embedded systems.
MICROPROCESSORS AND MICROSYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Alexandro Ortiz, Efrain Mendez, David Balderas, Pedro Ponce, Israel Macias, Arturo Molina
Summary: This study describes the implementation of metaheuristic optimization algorithms in hardware and compares five important algorithms. The results demonstrate the feasibility of NI FPGA hardware and reveal differences in device utilization and execution time among the algorithms.
APPLIED SOFT COMPUTING
(2021)
Article
Automation & Control Systems
Ning Wang, Jian-Nan Zhang, Hao Ni, Hong-Zhi Jia, Can Ding
Summary: This article proposes an improved fast terminal sliding-mode variable-structure control algorithm (FTSMC) to maximize power point tracking in thermoelectric generators. Experimental results show that the proposed algorithm can reach the nonlinear sliding mode more quickly and achieve a high conversion efficiency.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2022)
Article
Chemistry, Analytical
Juhyeon Park, Heoncheol Lee, Hyuck-Hoon Kwon, Yeji Hwang, Wonseok Choi
Summary: This paper proposes a parallelized particle swarm optimization technique using FPGA for real-time ballistic target tracking. Compared to conventional particle swarm optimization, the proposed method significantly reduces computation time.
Article
Computer Science, Hardware & Architecture
Fubo Ma
Summary: The research focuses on classifying and recognizing Indian classical dance videos using FPGA, with data collected from offline manual creation and online YouTube sources. Training and testing were conducted using different subjects and poses, resulting in a 90% recognition rate compared to other classification models.
MICROPROCESSORS AND MICROSYSTEMS
(2021)
Article
Engineering, Electrical & Electronic
Ching-Te Chiu, Yu-Chun Ding, Wei-Chen Lin, Wei-Jyun Chen, Shu-Yun Wu, Chao-Tsung Huang, Chun-Yeh Lin, Chia-Yu Chang, Meng-Jui Lee, Shimazu Tatsunori, Tsung Chen, Fan-Yi Lin, Yuan-Hao Huang
Summary: Face classification is important in various applications, but is challenging due to environmental variations. In this paper, we propose a Chaos LiDAR depth sensor for both indoor and outdoor applications, and design a face classification model based on RGB-D sensing. Experimental results show that our Chaos LiDAR sensor achieves higher classification accuracy compared to traditional sensors within a certain distance range.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS I-REGULAR PAPERS
(2022)
Article
Computer Science, Hardware & Architecture
Meizhen Zhang
Summary: Humans are social beings who rely on relationships based on their actions and abilities. The current psychological management system lacks balanced communication, leading to the proposal of a new method using FPGA and data mining to address relationship supervision challenges.
MICROPROCESSORS AND MICROSYSTEMS
(2021)
Article
Automation & Control Systems
Lahoucine Idkhajine, Eric Monmasson
Summary: This article proposes a new approach for designing embedded real-time simulators for power electronic converters. The approach utilizes dedicated coefficient varying transfer functions to approximate the voltage/current characteristics of each power switch. By employing parallel computing models and real measurements, high identification accuracy can be achieved. The use of field programmable gate array devices allows for low latencies and short simulation time steps.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2022)
Article
Computer Science, Hardware & Architecture
Luchuan Jiang
Summary: Virtual Reality (VR) technology enhances individuals' visual and motor skills, facilitating motor learning, particularly showing better results in swimming training. Research on swimming lesson systems using VR technology has the potential for cognitive and behavioral studies, allowing for comparative analysis between correct and incorrect swimming practices.
MICROPROCESSORS AND MICROSYSTEMS
(2021)
Article
Computer Science, Hardware & Architecture
Hongshu Bao, Xiang Yao
Summary: By utilizing augmented reality innovation, a basketball training reproduction framework is established, providing more targeted training by setting up a virtual reenactment model and capturing players' actual situations. With the use of Virtual Data Augmentation Technology (VDRT), players can quickly grasp key points of sports skills and significantly improve training efficiency.
MICROPROCESSORS AND MICROSYSTEMS
(2021)
Article
Automation & Control Systems
Bo-Lin Jian, Chia-Chuan Liu, Hao-Yang Lin, Her-Terng Yau
Summary: The purpose of this study was to design a vibrator for an ultrasonic tool holder and investigate how its shape, material, and design affect the tool holder's performance. Simulation and experiments were conducted to analyze resonance frequency, amplitude, and other factors. The results showed that different materials, bolt pretension forces, and tool sizes have an impact on the vibrator's performance.
INTERNATIONAL JOURNAL OF ADVANCED MANUFACTURING TECHNOLOGY
(2022)
Article
Engineering, Electrical & Electronic
Ping-Huan Kuo, Chia-Yu Lin, Po-Chien Luan, Her-Terng Yau
Summary: Cutting tool wear strongly affects machining processes. Traditional methods of measuring tool wear are time-consuming and not cost-effective. This study proposes using a CCD camera installed in the machine tool to estimate tool wear through programmed processing. However, this may affect the spatial configuration of the machine tool and require additional time for measurement. By combining an accelerometer sensor with an offline or online tool wear measurement system and using a training method, the time needed for measurement can be saved. The study develops a novel machine-learning algorithm using milling data and achieves high accuracy with a 1-D CNN model.
IEEE SENSORS JOURNAL
(2022)
Article
Engineering, Mechanical
Ping-Huan Kuo, Yung-Ruen Tseng, Po-Chien Luan, Her-Terng Yau
Summary: Chatter not only affects the surface quality of the workpiece and tool wear, but also increases production costs. Accurate detection of chatter signals is therefore necessary. Due to the nonlinear vibration nature of chatter during machining, different chatter characteristics are observed under different conditions. This research uses a machining learning method combined with a database and employs chaotic error mapping to accelerate data processing. With only 60 data points, an accuracy of 94.8% and precision of 99.62% can be achieved. Additionally, this research introduces the fractional-order (FO) convolutional neural network (FOCNN) for chatter detection, reducing trainable parameters by 42.3% compared to approximate training conditions while improving accuracy by 3.8%.
NONLINEAR DYNAMICS
(2023)
Article
Engineering, Multidisciplinary
Ping-Huan Kuo, Dian-Ying Cai, Po-Chien Luan, Her-Terng Yau
Summary: Cutter wear has a significant impact on machining quality, especially for high precision machining. This paper proposes a real-time machining status monitoring method using external sensor data. A tool wear forecast model is introduced, and multiple process parameters and sensor data are collected. The model is based on a Branched Neural Network and outperforms other algorithms in terms of prediction accuracy.
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL
(2023)
Article
Engineering, Electrical & Electronic
Ping-Huan Kuo, Yen-Wen Chen, Tung-Hsien Hsieh, Wen-Yuh Jywe, Her-Terng Yau
Summary: Considering the rapid development of technology, traditional manufacturing methods cannot achieve the required high accuracy in aerospace, national defense, and leading-edge engineering projects. Thermal displacement is a significant source of manufacturing errors, and it is difficult or even impossible to accurately correct such errors using traditional machining methods. This article proposes a machine learning method that can be easily implemented by non-professionals for high-accuracy error prediction. An optimized automatic logistic random generator time-varying acceleration coefficient particle swarm optimization (LRGTVAC-PSO) method is proposed to optimize a branch structured bidirectional gated recurrent unit (GRU) neural network. The proposed method achieves superior accuracy (with a three-axis average of 0.945) compared to other optimized algorithms evaluated in this study. The method not only accurately predicts thermal displacement but also autotunes the hyperparameters of machine learning algorithms.
IEEE SENSORS JOURNAL
(2023)
Article
Engineering, Electrical & Electronic
Chi-Yuan Lin, Shu-Cing Wu, Ping-Huan Kuo, Meng-Jun Huang, Song-Wei Hong, Her-Terng Yau
Summary: In recent years, industries have increased their demand for precision, automatic detection, and visualization interfaces. Machine tool operators install sensors on machine tools to obtain more precise measurements, but this results in complicated wire layouts. Wireless data transmission has emerged as a solution to this problem. However, while machine tool operators focus on optimizing sensing conditions, they often overlook information security. This study proposes a signal transmission encryption and decryption system for sensory toolholders during processing and compares two control methods for synchronization, finding that sliding mode control offers better synchronization and information security.
IEEE SENSORS JOURNAL
(2023)
Article
Engineering, Electrical & Electronic
Ping-Huan Kuo, Chieh-Hsiu Pao, En-Yi Chang, Her-Terng Yau
Summary: In this study, reinforcement learning was used to train humanoid robots to adapt to uneven terrains and automatically adjust their parameters for optimal gait pattern control. The results showed that proximal policy optimization (PPO), combining advantage actor-critic and trust region policy optimization, was the most suitable method. An improved version of PPO, called PPO2, was used in combination with data preprocessing methods such as wavelet transform and fuzzification, which improved the gait pattern control and balance of humanoid robots.
INTERNATIONAL JOURNAL OF OPTOMECHATRONICS
(2023)
Article
Engineering, Electrical & Electronic
Ping-Huan Kuo, Chia-Ho Lee, Her-Terng Yau
Summary: In the precision machining industry, thermal error is a common and difficult to control factor for machine tools. This study uses temperature sensors and an eddy current displacement meter to collect data for training models, which are then organized and normalized. Different learning models are used to predict the nonlinear factors that affect the errors, and the best two models with better predictive performance are identified for the pre-trained model of transfer learning. Retraining with Multilayer Perceptron (MLP) on these two models improves the predicted results, with an MAE value of 0.40, RMSE of 0.52625, and R-2 score of 0.99696.
INTERNATIONAL JOURNAL OF OPTOMECHATRONICS
(2023)
Article
Engineering, Multidisciplinary
Ping-Huan Kuo, Po-Chien Luan, Yung-Ruen Tseng, Her-Terng Yau
Summary: In this study, an effective procedure for chatter data preprocessing is proposed to improve neural network learning results from extremely low quantity data. By utilizing the characteristics of a chaotic attractor, the variability of chatter data can be minimized. A modified convolutional neural network and a deep convolutional generative adversarial net are used for improved chatter detection and classification. The proposed training strategy generates enough data to compensate for the lack of training data, providing a high-quality deep learning chatter detection model.
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL
(2023)
Article
Engineering, Electrical & Electronic
Chieh-Li Chen, Li-Hsuan Chen, Her-Terng Yau
Summary: This article analyzes the winding pattern and characteristics of filament winding cylinders for lightweight gas cylinder productions. A range of winding angles is adopted to prevent sliding between the filament material and the cylinder surface. The least composite filament can be determined by calculating the winding pattern using different winding angles. Motion planning for a four-axis filament winding machine can be carried out based on sequential contact points during the winding process.
Article
Engineering, Electrical & Electronic
Her-Terng Yau, Ping-Huan Kuo, Dian-Ying Cai, Chia-Yu Lin
Summary: Tool status testing is crucial for improving processing efficiency and quality. This study develops a tool wear monitoring model using sensor signals and machine learning. By preprocessing the signals and applying a neural network model, higher wear forecast accuracy and reduced training time cost can be achieved.
IEEE SENSORS JOURNAL
(2023)
Article
Energy & Fuels
Ping-Huan Kuo, Yung-Ruen Tseng, Po-Chien Luan, Her-Terng Yau
Summary: The Broad Transfer Learning Network (BTLN) model achieves similar prediction performance as a Multi-Layer Perceptron (MLP) model using only one-third of the parameters. It combines broad learning and transfer learning techniques, and improves performance by enhancing feature extraction and increasing training efficiency. The BTLN model shows a significant improvement of 18.5% in performance compared to common neural network models.
Article
Thermodynamics
Ping-Huan Kuo, Tzung-Lin Tu, Yen-Wen Chen, Wen-Yuh Jywe, Her-Terng Yau
Summary: Using AI algorithms, this study predicted the displacement of a cutting tool caused by thermal deformation based on data collected from machine tool experiments. Multiple machine learning models were constructed and evaluated for accuracy. The incorporation of transfer learning and model optimization was found to improve prediction accuracy and mitigate the negative effects of data collected at different times.
CASE STUDIES IN THERMAL ENGINEERING
(2023)
Article
Engineering, Multidisciplinary
Ping-Huan Kuo, Ssu-Chi Chen, Chia -Ho Lee, Po -Chien Luan, Her-Terng Yau
Summary: Many factors such as prolonged and high-intensity usage, tool-workpiece interaction, mechanical friction, and ambient temperatures can increase the temperature of a machine tool. This leads to spindle thermal displacement and reduced machining precision. To address this, an intelligent algorithm is used to predict the thermal displacement of the machine tool. The ensemble model of LSTM-SVM shows higher prediction performance compared to other machine learning algorithms.
CMES-COMPUTER MODELING IN ENGINEERING & SCIENCES
(2023)
Article
Engineering, Multidisciplinary
Ping-Huan Kuo, Ting-Chung Tseng, Po-Chien Luan, Her-Terng Yau
Summary: This study focuses on developing an effective method for predicting the remaining useful life of bearings to prevent machine damage and human accidents. By exploring neural network models and analyzing data, the study successfully predicts the remaining useful life with high accuracy. The proposed method is proven to be superior through a comparison with traditional models.
CMES-COMPUTER MODELING IN ENGINEERING & SCIENCES
(2022)
