• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2002년도 합동 추계학술대회 논문집 정보 및 제어부문
• /
• pp.163-168
• /
• 2002

The voting algorithm for action selection performs self-improvement by Reinforcement learning algorithm in the dynamic environment. The proposed voting algorithm improves the navigation of the robot by adapting the eligibility of the behaviors and determining the Command Set Generator (CGS). The Navigator that using a proposed voting algorithm corresponds to the CGS for giving the weight values and taking the reward values. It is necessary to decide which Command Set control the mobile robot at given time and to select among the candidate actions. The Command Set was learnt online by means as Q-learning. Action Selector compares Q-values of Navigator with Heterogeneous behaviors. Finally, real-world experimentation was carried out. Results show the good performance for the selection on command set as well as the convergence of Q-value.

• 한국군사과학기술학회지
• /
• 제10권2호
• /
• pp.21-30
• /
• 2007

A practical method to design the input shaping which generates control command is proposed in this paper, We suggest an experimental technique considering human operator's target tracking error to improve aiming accuracy which significantly affects hit probability. It is known that stabilization performance is one of the most important factors for ground combat vehicle system. In particular, stabilization error of the manual target tracking system mounted on moving vehicle directly affects hit probability. To reduce this error, we applied input command shaping method using preprocessing filtering and functional curve fitting. First of all, we construct the human operator model to consider effects of human operator on our system. Input shaping curve is divided into several regions to get rid of the above problems and to improve the system performance. At example design part, we chose three steps of functional command curve and determine the parameters of the function by the proposed design method. In order to verify the proposed design method, we carried out the experiments with real plant of a fighting vehicle.

• International Journal of Control, Automation, and Systems
• /
• 제1권1호
• /
• pp.43-53
• /
• 2003

A modified command shaping control to reduce residual vibrations at a target position and to limit the sway angle of the payload during traveling for container crane systems is investigated. When the maneuvering time is minimized, a large transient amplitude and steady state oscillations may occur inherently. Since a large swing of the payload during the transfer is dangerous, the control objective is to transfer a payload to the desired place as quickly as possible while limiting the swing angle of the payload during the transfer. The conventional shapers have been enhanced by adding one more constraint to limit intermediate sway angles of the payload. The developed method is shown to be more effective than other conventional shapers for prevention of an excessive transient sway. Computer simulation results are provided.

• 한국통신학회논문지
• /
• 제27권8C호
• /
• pp.771-781
• /
• 2002

기존의 PC 통신 서비스가 최근 인터넷에 기반을 둔 멀티미디어 웹 서비스로 빠르게 대체됨으로 인해 과거 많은 인력과 자금을 투자하여 개발한 유용한 데이터베이스와 PC 통신 환경 상의 많은 서비스가 사용자로부터 도외시되어 사장되는 문제가 발생하고 있다. 이에 본 연구에서는 과거 PC 통신 환경 상에서 제공되었던 텍스트 위주의 명령에 기초한 데이터베이스 서비스를 사용자의 정보 접근이 용이한 멀티미디어 웹에 기초한 데이터베이스 서비스로의 효율적 전환을 위한 구현방법 및 기술에 관해 연구하였다.

• 산업경영시스템학회지
• /
• 제28권4호
• /
• pp.134-141
• /
• 2005

Due to the complexity and stochastic nature of automated warehousing system, items are usually queued up at I/O point. We introduce a storage/retrieval policy : dual-command. We present quick approximations to queueing phenomena under these policies. It is assumed that the storage and retrieval arrival pattern follow the same poisson process. We also assumed that storage queue and retrieval queue being operated separately. The approximation attempts are performed under the proposed storage/retrieval policies after we derive S/R machine travel time distributions.

• 한국군사과학기술학회지
• /
• 제10권2호
• /
• pp.148-151
• /
• 2007

The CF(Counter-fire) is neutralizing enemy's all command control systems and fire support elements. It will weaken a battle continuous ability and an intension to fight. At the beginning of the CF is obtaining locations of targets using various detection assets. CF command center processes acquired target information and send it to attacking equipments. The targets are classified into two classes, preplanned target and target of opportunity The target of opportunity is potential threaten, so it needs to take a immediate and exact process for determining location of target of opportunity. This paper proposes the real-time processing algorithm for offensive weapons to strike target of opportunity, and presents the result of its performance.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1999년도 하계종합학술대회 논문집
• /
• pp.516-519
• /
• 1999

Gaze detection is to find out the position on a monitor screen where a user is looking at, using the image processing and computer vision technology, We developed a computer interface system using the gaze detection technology, This system enables a user to control the computer system without using their hands. So this system will help the handicapped to use a computer and is also useful for the man whose hands are busy doing another job, especially in tasks in factory. For the practical use, command signal like mouse clicking is necessary and we used eye winking to give this command signal to the system.

• Asian Pacific Journal of Cancer Prevention
• /
• 제13권9호
• /
• pp.4777-4781
• /
• 2012

Objective: To investigate the relationship between the efficacy and safety of different doses of thalidomide (Thal) plus dexamethasone (Dex) as the initial therapy in elderly patients with newly diagnosed multiple myeloma (MM). Methods: Clinical data of 28 elderly patients with newly diagnosed MM who underwent the TD regimen as the initial therapy were analyzed retrospectively. The patients were divided into two groups according to the maximal sustained dose of Thal: lower dose (group A) and higher dose (group B). The overall response rate (ORR), progression free survival (PFS), overall survival (OS), and adverse events (AES) were compared between the two groups. Results: A total of 28 patients were followed up with a median of 18 months. The ORR was 60.1%. The median response time and PFS were 2.0 and 17.0 months, respectively. The mean sustained dose of Thal in group B was significantly higher than group A (292.9 mg v 180.4 mg, P=0.01). There was no significantly difference in ORR (57.1% v 64.3%, P=1.00) and PFS (9.63months v 17.66 months, P=0.73) between groups A and B. During the follow up, only five patients died (<40%) and, therefore, median OS values were not available. It is estimated, however, that the mean survival time in the two groups was 35.6 and 33.4 months (P>0.05), respectively. All of the patients tolerated the treatment well. The incidence of AES in patients with a grading above 3 in group B was significantly higher than in group A (P=0.033). Conclusions: The TD regimen results in a high response rate and manageable AES as the initial therapy in elderly patients with MM. TD should be considered as the front line regimen for the treatment of elderly patients with MM in areas with financial constraints. The clinical response can be achieved at a low dose Thal with minimal toxicity.

• 한국소음진동공학회논문집
• /
• 제11권8호
• /
• pp.303-313
• /
• 2001

The identification of power flow in dynamically loaded structures Is essential in the analysis of structure-borne noise. However there are no general purpose tools to estimate powers flow. To make matters worse. It is very difficult to measure it. The power flow can be formulated in terms of balance forces(ELFORCE) at each element and velocities at the associated node obtained with MSC/NASTHAN. In this paper the procedure which is consist of the computations of the balance forces of al1 elements and the velocities at all nodes using MSC/NASTRAN. The calculations of the power f1ow at each element using PCL(PATRAN Command Language) and the Preparation of post -processes is set UP.

• 정보과학회 컴퓨팅의 실제 논문지
• /
• 제21권1호
• /
• pp.52-57
• /
• 2015

본 논문에서는 SSD의 TRIM 명령어 처리 성능을 LBA 범위의 특성 및 개수에 따라 분석하고, 그 결과를 바탕으로 Ext4 파일 시스템에서의 TRIM 명령어 처리 성능을 개선하였다. 대부분의 SSD에서 LBA 범위의 크기가 크고, 섹터 번호가 정렬 및 연속되어 있으며, 하나의 TRIM 명령어로 다수의 LBA 범위 정보를 전달할수록 TRIM 명령어의 처리 성능이 증가하는 것으로 나타났다. 한편, 기존의 Ext4 파일 시스템은 이와 같이 다수의 LBA 범위 정보를 전달할 때 그 성능이 더 높음에도 불구하고, 하나의 TRIM 명령어 당 하나의 LBA 범위 정보만 전달하는 문제점을 가지고 있다. 이를 해결하기 위해 하나의 TRIM 명령어로 최대 64개의 LBA 범위 정보를 전달하도록 Ext4 파일 시스템을 개선하였고, Filebench를 이용한 성능 평가에서 파일 삭제 성능이 최대 35% 증가하는 것을 확인하였다.