• 한국항공우주학회지
• /
• 제38권5호
• /
• pp.427-436
• /
• 2010

본 논문은 다수의 목표물과 다수의 임무가 존재하는 상황에서의 다수 무인항공기의 임무할당 문제를 다룬다. 다수 무인항공기의 임무할당 문제는 순회 세일즈맨 문제, 차량 라우팅 문제와 같은 조합최적화 문제의 일종으로 NP-hard의 계산 복잡도를 가지고 있다. 이런 성격의 문제는 문제의 크기가 커질수록 계산시간이 급격히 증가하는 특징을 지니기 때문에 문제를 효율적으로 풀기 위해서 근사화 방법 또는 발견적인 방법을 사용한다. 본 연구에서는 임무할당 문제를 혼합정수 선형계획 문제로 정식화하고, 혼합정수 선형계획법과 유전 알고리듬으로 해를 구하였다. 다수의 목표물, 다수의 임무, 장애물이 존재하는 환경에 대한 수치 시뮬레이션을 수행하여 각 방법의 최적성과 효율성에 대해 검토하였다.

• 한국정보과학회논문지:컴퓨팅의 실제 및 레터
• /
• 제15권4호
• /
• pp.304-308
• /
• 2009

본 논문은 실시간 임베디드 OS인 TMO-eCos의 데드라인 기반 CPU 저전력 관리 기법을 다루고 있다. 해당 저전력 관리 기법은 경성 실시간 시스템인 TMO 시스템을 위한 태스크 순차화 기법에서 도출된 스케줄링 시나리오를 사용한다. 본 연구팀에서 개발한 스케줄링 사전 분석기는 주기적으로 동작하는 태스크의 주기, 데드라인, WCET를 기반으로 오프라인 분석을 실시한다. 최종적으로 TMO-eCos 커널은 CPU의 전력 소모를 줄이기 위하여 주기적인 태스크의 데드라인을 위반하지 않는 범위에서 CPU의 속도를 조절하여 시스템에서 사용하는 소비전력은 줄이게 된다. 본 논문은 이와 같은 과정과 실제 실험결과를 기술한다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권12호
• /
• pp.4081-4098
• /
• 2022

With the aim of tackling the contradiction between computation intensive industrial applications and resource-weak Edge Devices (EDs) in Industrial Internet of Things (IIoT), a novel computation task offloading scheme in SDIN-enabled MEC based IIoT is proposed in this paper. With the aim of reducing the task accomplished latency and energy consumption of EDs, a joint optimization method is proposed for optimizing the local CPU-cycle frequency, offloading decision, and wireless and computation resources allocation jointly. Based on the optimization, the task offloading problem is formulated into a Mixed Integer Nonlinear Programming (MINLP) problem which is a large-scale NP-hard problem. In order to solve this problem in an accessible time complexity, a sub-optimal algorithm GPCOA, which is based on hybrid evolutionary computation, is proposed. Outcomes of emulation revel that the proposed method outperforms other baseline methods, and the optimization result shows that the latency-related weight is efficient for reducing the task execution delay and improving the energy efficiency.

• 정보처리학회논문지:컴퓨터 및 통신 시스템
• /
• 제3권6호
• /
• pp.173-178
• /
• 2014

제시된 스케줄링 기법은 병렬처리 기법을 활용하여 실시간 작업들의 데드라인 제약과 결함 포용 제약을 만족하면서 멀티코어 프로세서의 에너지 소모 효율성을 향상시켰다. 최소 에너지 소모량 스케줄을 찾는 것은 NP-hard 문제이므로, 제시된 기법은 다항식의 시간 내에 최소 에너지 소모량에 근접하는 스케줄을 찾는다. 제시된 기법은 연관된 최신 기법과 비교하여 높은 병렬처리 속도는 물론 낮은 병렬처리 속도에서도 에너지 소모량이 현격하게 낮았으며, 에너지 소모량을 최대 86% 줄였다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 ITC-CSCC -1
• /
• pp.443-445
• /
• 2002

In this paper, we present a novel efficient semaphore implementation scheme which diminishes completion time of high priority tasks and improves reliability of a system. The real-time system is constrained to complete their tasks in time. Especially, the task of a hard real-time system must meet its deadline under unfavorable conditions. In this paper, the number and sort of the locked semaphores, when an event occurred, decide whether the context switch should occur or not, so higher priority tasks diminish in their completion time. The experimental results show that the proposed method gives performance improvements in finish time of high priority tasks of about 11% over the Zuberi.

• Journal of Biosystems Engineering
• /
• 제38권2호
• /
• pp.163-170
• /
• 2013

Purpose: The purpose of this study is to provide analysis of physiological, biomechanical responses occurring from the operation to lifting or twist lifting task appears frequently in agricultural work. Methods: This study investigated the changes of physiological factors such as heart rate, heart rate variability (HRV) and biomechanical factors such as physical activity and kinetic analysis in the task of twisting at the waist while lifting. Results: Heart rates changed significantly with the workload. The result indicated that the workload of 2 kg was light intensity work, and the workload of 12 kg was hard intensity work. Physical activity increased as the workload increased both on wrist and waist. Besides, stress index of the worker increased with the workload. Dynamic load to herniated discs was analyzed using inertial sensor, and the angular acceleration and torque increased with the workload. The proposed measurement system can measure the recipient's physiological and physical signals in real-time and analyzed 3-dimensionally according to the variety of work load. Conclusions: The system we propose will be a new method to measure agricultural workers' multi-dimensional signals and analyze various farming tasks.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권10호
• /
• pp.4080-4097
• /
• 2020

In recent years, convolutional neural network (CNN) has become the primary method for face detection. But its shortcomings are obvious, such as expensive calculation, heavy model, etc. This makes CNN difficult to use on the mobile devices which have limited computing and storage capabilities. Therefore, the design of lightweight CNN for face detection is becoming more and more important with the popularity of smartphones and mobile Internet. Based on the CPU real-time face detector FaceBoxes, we propose a multi-task lightweight face detector, which has low computing cost and higher detection precision. First, to improve the detection capability, the squeeze and excitation modules are used to extract attention between channels. Then, the textual and semantic information are extracted by shallow networks and deep networks respectively to get rich features. Finally, the landmark detection module is used to improve the detection performance for small faces and provide landmark data for face alignment. Experiments on AFW, FDDB, PASCAL, and WIDER FACE datasets show that our algorithm has achieved significant improvement in the mean average precision. Especially, on the WIDER FACE hard validation set, our algorithm outperforms the mean average precision of FaceBoxes by 7.2%. For VGA-resolution images, the running speed of our algorithm can reach 23FPS on a CPU device.

• 기본간호학회지
• /
• 제4권2호
• /
• pp.193-216
• /
• 1997

The purpose of this research was to understand the structure of the experiences of turnover among hospital nurses. The research question was 'What is the structure of the experiences of turnover among hospital nurses?' The sample consisted of 16 hospital nurses who experienced one or more turnover. The unstructured interview were carried out from July 15. through August 31., 1997. Intervewed by audio-recording and analyzed by Van Kaam's method. There were 361 descriptive expressions and priority classifications. The result summerized as 107 common elements, 38 syntheses of hypothetical definitions and 6 identifications of the structural definition. The structural definitions and hypothetical definitiones were as follows ; 1. There were different views on the turnover intention. Hard task, dissatisfaction of initial expectancy, uncomfortable relationship with doctors, inappriate reward, uncomfortable relationship with workers of other parts, uncomfortable relationship with patients, unreasonable promotion, uncomfortable relationship with co-worker, helplessness, poor environment of working conditions. 2. Motivating factor of turnover can be varied : self development, problem of social support, boring task, problem of relation-ship with the senior, unfair promotion, night duty. 3. Social supportive factors were co-worker support, senior support, self control, family support, time support. 4. There were different views on the job satisfaction. Task, relationship with patient, recognition, professionalism, promotion, working environment, relationship with co-worker, reward.

• 한국컴퓨터정보학회논문지
• /
• 제24권3호
• /
• pp.95-104
• /
• 2019

Crowdsourcing is growing its interests in designing where various designers work independently to a given design task. Recent researchers discovered that collaboration by sharing designs among designers helps to produce high-quality designs. However, design task may still be hard even with that collaboration in case the requirements are not well-defined. Most customers sometimes do not know what they really want and do not know how to clearly define the requirements. Consequently, the lack of requirements creates issues on designers, such as spending much time and effort on collecting requirements alone or from the customers. The designers even end up missing important necessities to complete their tasks. To address this issue, we proposed a collaborative requirements elicitation method that supports designers who are working on the same task. We developed CREFD (Collaborative Requirements Elicitation For Designers and Developers) tool to enable designers collaboratively provide requirements, identify dependencies, add annotations and votes to the provided requirements. We performed the hypothetical and empirical evaluations to test and compare the proposed method with one of the existing elicitation methods, the results show that the proposed method helps in collecting accepted and well-organized requirements better than individual requirements elicitation.

• 정보처리학회논문지A
• /
• 제18A권6호
• /
• pp.255-264
• /
• 2011

센서 네트워크는 빛, 소리, 온도, 움직임 같은 물리적 데이터를 센서 노드에서 감지하고 측정하여 중앙으로 전달하고 처리하는 구조를 가진 네트워크이다. 센서 네트워크는 여러 분야에서 활용할 수 있는 기술이다. 센서 노드가 외부에서 채취한 데이터를 실시간으로 사용자에게 전달하는 것은 매우 중요하다. 센서 네트워크의 핵심은 센서 노드인 하드웨어 플랫폼과 노드에 들어가는 초소형 운영체제라고 할 수 있다. UC 버클리에서 개발된 TinyOS는 센서 노드에서 동작하는 운영체제 중 가장 많이 사용되고 있다. TinyOS는 Event-driven 방식이며 Component 기반의 센서 네트워크 운영체제이다. 기본적으로 비선점 방식의 스케줄러를 사용한다. 만약 급한 작업이 수행되어야 하는 시점에서 다른 태스크가 수행 중에 있다면 수행 중인 태스크가 완료할 때까지 기다려야 한다. 이러한 특성으로 인해 TinyOS에서 정해진 시간안에 자신의 작업을 끝낸다고 보장하기 어렵다. 최근 연구에서 TinyOS의 빠른 반응성을 위해 Priority Level Scheduler라는 선점 기능이 제안되었다. 이것은 제한적으로 5개의 우선순위를 만들어 높은 우선순위가 낮은 우선순위를 선점할 수 있게 한다. 여기서 본 논문은 TinyOS의 실시간성을 보장함과 더불어 사용자 태스크의 평균 응답시간을 줄이고자 Priority Level Scheduler에 실시간 스케줄러인 EDF(Earliest Deadline First)를 적용한 선점형 EDF 스케줄링 방식을 제안하고자 한다.