http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
A Comparison of Using Probabilistic Motion Models for Mobile Robot Pose Estimation
T.T. Quyen Bui,Keum-Shik Hong 제어로봇시스템학회 2009 제어로봇시스템학회 국제학술대회 논문집 Vol.2009 No.8
This paper addresses a robot pose estimation problem. The data received from the inertial sensors are converted to a diverse odometry, after that the current robot pose is estimated by utilizing the different probabilistic motion models and estimate algorithms. In addition, the experimental evaluations are performed in indoor environment and conclusions are given.
Sonar-based obstacle avoidance using region partition scheme
Bui, T.T.Quyen,Hong, Keum-Shik 대한기계학회 2010 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.24 No.1
This paper addresses an obstacle avoidance problem for a mobile robot in indoor environment. The collision avoidance algorithm utilizes a region partition scheme to analyze the three forward sensory regions of the robot, then to calculate the minimum distances from the center of mass of the robot to the obstacles, respectively, and ultimately to find a navigable region. The supervisor inside the controller receives a series of obstacle avoidance behaviors and makes a decision that allows the robot to successfully navigate in a cluttered environment. In addition, experimental results to illustrate the proposed obstacle avoidance algorithm for the mobile robot in navigation are also presented.
A Glove-Based Gesture Recognition System for Vietnamese Sign Language
Lam T. Phi,Hung D. Nguyen,T.T. Quyen Bui,Thang T. Vu 제어로봇시스템학회 2015 제어로봇시스템학회 국제학술대회 논문집 Vol.2015 No.10
In this paper, we address a development of a glove-based gesture recognition system for Vietnamese sign language. A sensor glove is attached ten flex sensors and one accelerometer. Here, flex sensors are used for sensing the curvature of fingers and the accelerometer is used in detecting a movement of a hand. Depending on the hand’s postures, i.e., vertical, horizontal, and movement, sign language of letters of the Vietnamese alphabets can be divided to category 1, 2, and 3, respectively. Firstly, the hand’s posture is recognized. Next, if the hand’s posture belongs to either category 1 or category 2, a matching algorithm is used to detect a letter. If the posture belongs to category 3, a dynamic time warping algorithm is applied. The use of our system in recognizing Vietnamese sign language is illustrated. In addition, experimental results are provided.
Hoang-Anh T. Nguyen,Huy-Hoang Tran,Thang T. Vu,T.T. Quyen Bui 제어로봇시스템학회 2016 제어로봇시스템학회 국제학술대회 논문집 Vol.2016 No.10
Mental state classification is an important step for realizing a control system based on electroencephalography (EEG) signals which could benefit a lot of paralyzed people including the locked-in or Amyotrophic Lateral Sclerosis. Considering that EEG signals are nonstationary and often contaminated by various types of artifacts, classifying thoughts into correct mental states is not a trivial problem. In this work, our contribution is that we present and realize a novel model which integrates different techniques: Independent component analysis (ICA), relative wavelet energy, and support vector machine (SVM) for the same task. We applied our model to classify thoughts in two types of experiment whether with two or three mental states. The experimental results show that the presented model outperforms other models using Artificial Neural Network, K-Nearest Neighbors, etc.
Extended Airtime Fair Scheduling Method for a Zigbee Network
Van-Tu Chau,T.T. Quyen Bui,Anh-Tuan Do,Quang-Vinh Thai 제어로봇시스템학회 2014 제어로봇시스템학회 국제학술대회 논문집 Vol.2014 No.10
In a Zigbee network, when many nodes simultaneously send data packets to a Zigbee coordinator through routes (branches), the Zigbee coordinator cannot deal with all signals. Interference and noises are existed during data transmission. This problem is still one of challenges in the network communication. In this paper, we propose an extended airtime fair algorithm in which throughput of a Zigbee network is maximized and fractions of airtime are taken into account. We adapt a utility function that is proposed by Kelly (1997) in maximizing the network throughput; nevertheless, throughput of each branch is multiplied by a fraction of airtime. Fractions of airtime play a role in creating a balance of throughputs of branches: the throughput and fairness of each branch is guaranteed. Moreover, the usefulness of our proposed algorithm for a Zigbee network is presented; simulation results show that the throughput and fairness are better than those of a case of not using airtime fair.
Design of an Advanced Wearable Sensor Platform for Multi Applications
Anh-Tuan Nguyen,Tung Hoang,Quang-Vinh Thai,T.T. Quyen Bui 제어로봇시스템학회 2015 제어로봇시스템학회 국제학술대회 논문집 Vol.2015 No.10
In the emerging Internet-of-Things (IoT), wireless sensors are the vital mediators between the physical world and the cyber space. In many telemonitoring and interactive applications, high-rate data streams up to MB/s may need to be transported from the wireless sensors to the cloud computing servers over the Internet and processed by the remote servers in quasi-real time fashion. These applications generally demand substantial reduction of communication bandwidth, response latency and power consumption of these Internet-based cyber-physical systems. To tackle these demands for efficient use of communication, computing and power resources, we developed an advanced wearable sensor platform (AWSP) which integrated a powerful system-on-chip (SoC) processor, a smart power management unit, a highly accurate real-time clock and multi-function peripherals into a miniature module. This AWSP designed to function is capable of performing sophisticated data pre-processing including real-time artifact and noise removal, data compression, and even feature extraction before uploading the data to other devices like mobile phone, computer, etc. Furthermore, with the installation of embedded real-time Linux operating system, this sensor provides a familiar and powerful software development environment for system developers to build their computation-intensive real-time applications. This paper presents the design and development process.
Development and Control Strategies of an Unmanned Autonomous Forklift
홍금식(Keum-Shik Hong),김민환(Min-Hwan Kim),이석(Suk Lee),김성신(Seong-Sin Kim),홍봉희(Bong Hee Hong),위디오트리앗모 아우기(A. Widyotriatmo),브이티탄 규엔(Quyen T. T. Bui),이정훈(Jung Hun Lee),탐바 투아 아구스티누스(Tua A. Tamba) 대한기계학회 2008 대한기계학회 춘추학술대회 Vol.2008 No.5
This paper presents the control strategies for the development of an unmanned autonomous forklift. A control architecture using several sensors (vision, ultrasonic sensor, laser range finder, etc.) and a PC controller is presented. An ultrasonic satellite and gyro are used for the localization, and a vision system is utilized for obstacle detection and the measurement of pallet position. The CAN communication network is utilized to provide the drive-by-wire technology for the unmanned autonomous forklift.