• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2000년도 춘계학술대회 학술발표 논문집
• /
• pp.228-231
• /
• 2000

A hybrid fuzzy controller by means of the genetic algorithms is presented. The control input for the system in the HFC is a convex combination of the FLC's output in transient state and PlD's output in steady state by a fuzzy variable. The HFC combined a PID controller with a fuzzy controller concurrently produces the better output performance than any other controller. A auto-tuning algorithms is presented to automatically improve the performance of hybrid fuzzy controller using genetic algorithms. The algorithms estimates automatical Iy the optimal values of scaling factors, PID parameters and membership function parameters of fuzzy control rules. Especially, in order to auto-tune scaling factors and PID parameters of HFC using GA three kinds of estimation modes are effectively utilized. The HFCs are applied to the second process with time-delay. Computer simulations are conducted at step input and the performances of systems are evaluated and also discussed in ITAE(Integral of the Time multiplied by the Absolute value of Error ) and other ways.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 1998년도 The Third Asian Fuzzy Systems Symposium
• /
• pp.51-56
• /
• 1998

In this paper, we analyze the effects of scaling factors on the performance of a fuzzy logic controller(FLC). The quantitative relation between input and output variables of FLC is obtained by using a qualsi-linear fuzzy model, and an approximate transfer function of FLC is dervied from the comparison of it with the conventional PID controller. Then we analyze in detail the effects of scaling factor using this approximate transfer function and root locus method. Also we suggest an on-line tuning method for scaling factors which employs an sample performance function and a variable reference for tuning index.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1999년도 추계학술대회 논문집 학회본부 B
• /
• pp.521-523
• /
• 1999

In this paper, we propose the hybrid fuzzy controller(HFC) and the hybrid auto-tuning algorithm. The proposed HFC combined a PID controller with a fuzzy controller concurrently produces the better output performance such as sensitivity improvement in steady state and robustness in transient state than any other controller. In addition, a hybrid auto-tuning algorithm which consists of genetic algorithm and complex algorithm to automatically generate weighting factor, scaling factors and PID control gains optimizes the output of HFC. As an typical example of non-linear system in control theory an inverted pendulum will be controlled by the suggested HFC and illustrated the performance and applicability of this proposed method by simulation.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2015년도 추계학술발표대회
• /
• pp.197-198
• /
• 2015

IT 기술의 진보적 발전에 따라 클라우드 컴퓨팅 분야 연구들이 활발히 진행되고 있다. 클라우드 컴퓨팅은 가상화 기술을 이용하여 크게 인프라, 플랫폼, 소프트웨어 관점으로 나뉘어 사용자에게 다양한 서비스를 제공한다. 가상화 기술 중에 Desktop Storage Virtualization (DSV)은 분산된 레거시 데스크탑으로 구성되어 있기 때문에 비가용 상태 시간별 클러스터링 및 사용자 요청에 따른 자동 확장이 매우 중요시된다. 본 논문에서는 GPU의 many-core를 이용하여 분산된 데스크탑의 성능 상태 분석 및 자동 확장을 위해 스레드별로 호스트를 매핑하고 병렬적으로 처리하는 Rapid Auto Scaling Mechanism (RASM)을 제안한다.

• 대한전자공학회논문지TC
• /
• 제44권3호
• /
• pp.33-40
• /
• 2007

본 논문에서는 최근 새롭게 발견된 low-rate TCP (LRT) 공격과 이 공격을 감지하기 위한 DTW (Dynamic Time Warping) 알고리즘을 분석하고 공격 검출에 대한 성능 향상을 위한 스케일링 기반 DTW (Scaling based DTW; S-DTW) 알고리즘을 소개한다. Low-rate TCP 공격은 대용량 트래픽을 사용한 기존 서비스 거부 공격과는 다르게 공격 트래픽의 평균 트래픽 양이 적어서 기존 DoS 공격에 대한 감지 방식으로는 검출되지 않는다. 그러나 LRT 공격은 주기적이고 짧은 버스트 트래픽으로 TCP 연결의 최소 재전송 타임아웃 (Retransmission Timeout; RTO)에 대한 취약성을 공격하기 때문에 패턴 매칭으로 공격 감지가 가능하다. 기존 메커니즘에 의한 감지 기법은 공격 패턴의 입력 샘플 템플릿을 기준으로 입력 트래픽이 정상 트래픽인지 또는 공격 트래픽인지를 판별한다. 이 과정에서 입력 트래픽의 특성에 따라서 DTW 알고리즘은 정상 트래픽을 공격 트래픽으로 오판하는 문제점을 갖는다. 따라서 본 논문에서는 이러한 오판을 줄이기 위하여 기존 DTW 알고리즘의 전처리 과정인 자기상관 (auto-correlation) 처리를 분석하여 오판을 규명한다. 또한 스케일링 기반으로 자기상관 처리 결과를 수정하여 공격 트래픽과 정상 트래픽의 특성의 차이를 증가시킴으로써 DTW 알고리즘에 의한 공격 감지 능력을 향상시킨다 마지막으로 다양한 스케일링 방식과 표준편차에 의한 트래픽 분석 방법도 논의된다.

• International Journal of Contents
• /
• 제13권2호
• /
• pp.29-34
• /
• 2017

Cloud computing is becoming an effective and efficient way of computing resources and computing service integration. Through centralized management of resources and services, cloud computing delivers hosted services over the internet, such that access to shared hardware, software, applications, information, and all resources is elastically provided to the consumer on-demand. The main enabling technology for cloud computing is virtualization. Virtualization software creates a temporarily simulated or extended version of computing and network resources. The objectives of virtualization are as follows: first, to fully utilize the shared resources by applying partitioning and time-sharing; second, to centralize resource management; third, to enhance cloud data center agility and provide the required scalability and elasticity for on-demand capabilities; fourth, to improve testing and running software diagnostics on different operating platforms; and fifth, to improve the portability of applications and workload migration capabilities. One of the key features of cloud computing is elasticity. It enables users to create and remove virtual computing resources dynamically according to the changing demand, but it is not easy to make a decision regarding the right amount of resources. Indeed, proper provisioning of the resources to applications is an important issue in IaaS cloud computing. Most web applications encounter large and fluctuating task requests. In predictable situations, the resources can be provisioned in advance through capacity planning techniques. But in case of unplanned and spike requests, it would be desirable to automatically scale the resources, called auto-scaling, which adjusts the resources allocated to applications based on its need at any given time. This would free the user from the burden of deciding how many resources are necessary each time. In this work, we propose an analytical and efficient VM-level scaling scheme by modeling each VM in a data center as an M/M/1 processor sharing queue. Our proposed VM-level scaling scheme is validated via a numerical experiment.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1999년도 추계학술대회 논문집 학회본부 B
• /
• pp.588-590
• /
• 1999

In this paper the design method of a fuzzy logic controller with a genetic algorithm is proposed. Fuzzy logic controller is based on linguistic descriptions(in the form of fuzzy IF-THEN rules) from human experts. The auto-tuning method is presented to automatically improve the output performance of controller utilizing the genetic algorithm. The GA algorithm estimates automatically the optimal values of scaling factors and membership function parameters of fuzzy control rules. Controllers are applied to the processes with time-delay and the DC servo motor. Computer simulations are conducted at the step input and the output performances are evaluated in the ITAE.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1997년도 하계학술대회 논문집 B
• /
• pp.728-731
• /
• 1997

This paper proposes the systematic auto-tuning method for fuzzy controller using genetic algorithm(GA). In general, the design of fuzzy logic controller has difficulties in the acquisition of expert's knowledge and relies to a great extent on heuristic knowledge which, in many cases, cannot be objectively justified. So, the performance of the controller can be degraded in the case of plant parameter variations or unpredictable incident which the designer may have ignored. Proposed genetic algorithm searches the optimal rule structure, parameters of membership functions and scaling factors simultaneously and automatically by a new genetic coding format. Inverted pendrum system is provided to show the advantages of the proposed method.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1996년도 하계학술대회 논문집 B
• /
• pp.1034-1036
• /
• 1996

In this paper, an adaptive fuzzy logic controller is presented for auto-tunning of the scaling factors by using learning capability of neural networks. The proposed scheme consists of the FLC which includes the PI-type FLC and PD-type FLC in parallel form and the neural network which learns scale factors of FLC. Computer simulations were performed to illustrate the effectiveness of a proposed scheme. A proposed FLC controller was applied to the second order system and velocity control of the brushless DC motors. For the design of the FLC, tracking error, change of error, and acceleration error are selected as input variables of the FLC and three seal e factors were used in the parallel-type FLC. This scheme can be used to reduce the difficulty in the selection of the scale factors.

• 자동차안전학회지
• /
• 제11권1호
• /
• pp.30-35
• /
• 2019

KIDAS (Korean In-Depth Accident Study) is a data structure of accident investigation type, vehicle breakage and human injury database. A consortium of research institutes, universities, and medical institutions has been established and operated. KIDAS has the strongest difference from the TAAS (Traffic Accident Analysis System), which is the data of the National Police Agency, that it can grasp the injury information of passengers. In this study, the mean age and weight of the most frequent accident types in the KIDAS accident statistics were calculated to determine the degree of injury according to gender. Through the MADYMO analysis, it is aimed to grasp the difference of dummy injury using commercial dummy models and scaling models are currently used.