• Journal of Korea Society of Industrial Information Systems
• /
• v.20 no.3
• /
• pp.51-60
• /
• 2015

Traceability targets provision of information to stakeholders required for analyzing impacts among artifacts due to changes. Unlike single product development, in software product line developing the family of products the complexity of maintaining and managing traceability between two life cycles, domain and application engineering is so high. Accordingly, variability traceability management approach centred on orthogonal variability model that manages variability separated from development artifacts has been conceptually proposed, but its efficiency has not verified yet. This paper verifies whether orthogonal variability model based traceability can provide required traceability through an example. As the results, the OVM-based variability tracing method supports well to narrow down artifacts affected by the changes. However, the method does not support tracing the exact artifacts or exact part of an artifact affected by the change.

• Journal of KIISE:Software and Applications
• /
• v.34 no.4
• /
• pp.328-341
• /
• 2007

When the decision to initiate a software product line has been taken, the first step is the domain analysis describing the variability in the requirements, the second important step is the definition of a domain architecture that captures the overall structure of a series of closely related products. A domain architecture can be a core asset in product line by describing the commonalities and variabilities of the products contained in the software product line. The variabilities, which are identified at each phase of the core assets development, are diverse in the level of abstraction. Therefore, it is important to clearly define, systematically identify, and explicitly represent variability at the architectural level. However, it is difficult to identify and represent the variability which should be considered at the architecture level, because these may be appeared in architecture elements and in architecture configuration. In this paper, we suggest a method of developing domain architecture as a core asset in product line where commonality and variability are explicitly considered. First of all, we will describe a domain architecture metamodel that can explicitly define commonality and variability concepts by extending the Object Management Group's ($OMG^{TM}$ Reusable Asset Specification eRAS) model. Using the domain architecture metamodel, architecture elements are defined and the variations that can be identified at the architecture level are classified into two types in according th abstract level. Additionally, we describe a domain architecture where commonality and variability are explicitly considered on basis of this metamodel.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.18 no.1
• /
• pp.1-6
• /
• 2008

In this paper, we present the robust fuzzy algorithm for variable speed control of wind turbines. Generally, the plants of wind turbines are consisted of complex nonlinearities, and the parameters of variable speed of wind turbines are represented as uncertain terms. For solving these complexity, we propose the robust fuzzy algorithm. At first, the exact fuzzy modeling are performed for variable speed of wind turbines. Next, we design the fuzzy controller for reanalyzed T-S fuzzy model of the wind turbines, then, we prove the stability of the plant through the Lyapunov stability theorem. At last, an example is included for visualizing the efficiency of the proposed technique.

• Journal of KIISE:Software and Applications
• /
• v.30 no.5_6
• /
• pp.414-429
• /
• 2003

Component-based development(CBD) has been generalized in industry to master the complexity and reduce the development cost and time. However, current CBD practice is developing the component which is dependent on single application[l][2]. Therefore component variability is emphasized to reuse the component in many family members in a domain[8]. However, components are developed for the reason of replaceability rather than the reusability which is the main purpose of the component due to the insufficiency of the study of component variability definition and type[3]. In this paper, we formally specify the component variability reflecting the characteristics of the component to increase the component reusability. We define the logic variability which was recognized as the existing component variability and we propose all types of variability existing in the component by suggesting three more variability types. And we propose the component variability scope which makes us estimate and verify the number of cases of the variability when we customize the component. We propose these component variability types and scope through formal specification. By applying these techniques in developing components, we can develop high quality components reusable in many family members.

• Journal of KIISE:Software and Applications
• /
• v.27 no.9
• /
• pp.920-930
• /
• 2000

컴포넌트 기반의 소프트웨어 개발이 소프트웨어 복잡성, 비용, 그리고 품질을 해결하기 위한 새로운 대안으로 소개되고 있다. COM, Enterprise JavaBeans, CORBA 컴포넌트 모델등과 같은 다양한 컴포넌트 아키텍쳐들이 소개되고 있으며 컴포넌트 기반의 소프트웨어 개발 방법론과 여러 CASE 도구들이 이를 지원하고 있다.[1,2,3,4]. 그러나 현재 컴포넌트를 구현할 수 있는 기술은 제시되어 있지만 컴포넌트를 모델링하는 기법들에 대한 연구는 미약한 상태이다. 본 논문에서는 도메인 분석에서 공통성과 가변성 추출 및 클러스터링 기법을 이용한 컴포넌트를 분석하는 기법을 제시한다. 즉 컴포넌트 추출 기법, 컴포넌트의 핫스팟(또는 가변성)표현 기법, 컴포넌트 요구사항 정의 기법 등을 제시한다. 컴포넌트 개발에 있어서 이러한 모델링 기법을 적용함으로써 컴포넌트를 효율적으로 개발할 수 있을 뿐만 아니라 재사용성이 높은 고품질의 컴포넌트 개발을 지원할 수 있다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2006.10b
• /
• pp.254-259
• /
• 2006

전통적인 인공지능 계획방식은 완전한 월드 상태모델과 시스템 동작모델에 기초하여 처음부터 자동으로 작업계획을 생성하려는 접근방식이다. 그러나 지능로봇제어와 같이 불확실성과 가변성이 높은 실 세계 응용분야에서 이와 같은 전통적인 인공지능 계획방식은 효과를 얻기 어렵다. 반면에 많은 실 세계 응용분야에서는 그 분야에서 이미 잘 알려져 있는 작업 영역지식이나 제어지식들이 존재하며, 이들을 효과적으로 이용하는 것이 매우 중요하다. 이러한 방법 중의 하나로서 복잡도가 높은 작업계획을 전문가가 직접 편집해서 입력하는 방식이 널리 쓰인다. 기본 동작모델과는 달리, 일반적으로 작업계획 표현언어는 복잡한 제어구조를 포함하는 하나의 작업 프로세스로 계획을 표현한다. 따라서 이러한 복잡한 절차적 지식인 작업계획을 편집하고 검증하기 위해서는 편리한 모델링 도구의 개발이 필요하다. 본 연구에서는 PRS 계열의 작업계획을 비주얼 환경에서 편집할 수 있고, 가상 시뮬레이션 기능과 작업 계획기와의 연동 기능을 갖춘 PKM시스템의 설계와 구현에 대해 설명한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.10b
• /
• pp.39-41
• /
• 2000

S/W 산업 시대의 국방정보시스템은 대규모화되고 높은 실시간성과 정확성의 요구로 S/W의 복잡성이 증가하고 있으며, 요구되는 기능과 성능이 매우 다양해지고, 사용자의 요구사항도 매우 가변적인 방향으로 발전하고 있다. 본 논문에서는 국방정보시스템의 특성을 명확히 파악하고, 특성에 따라 국방정보시스템 S/W 개발비용 산정에 영향을 미치는 영향요소를 식별한 후, 기존의 '한소협' 모델을 바탕으로 S/W 개발비용 보정계수로 적용할 영향요소와 기능점수를 산정하기 위한 기술적 복잡도 요소로 재분류하였으며, 기존의 '한소협' 모델의 절차와 각종 보정계수의 적용방법을 국방정보시스템에 적합하도록 수정 보완하는 방식으로 국방정보시스템의 S/W 개발비용 산정 지원용 전문가시스템의 도메인 지식을 개발하였으며, 이러한 도메인 지식 개발은 궁극적으로 완벽한 전문가시스템 개발에 밑거름이 될 것이다.

• The KIPS Transactions:PartD
• /
• v.13D no.7 s.110
• /
• pp.947-958
• /
• 2006

In order to increase extensibility and reusability of business components, the variable things need to be analyzed from the analysis phase and identified as components. In this paper, we propose a rule-based analysis pattern, which can effectively extract object-based main concepts from a variable business process in the analysis phase and identify a variable business component by applying the pattern to the UML Components development process. It can make analysis artifacts consistent and readable for analysts with different level of knowledge and experience to apply the pattern to analysis of rule-based variable business processes. And also, variable business components can be easily identified by applying the pattern to the UML Components development process. In order to prove the feasibility of the pattern, we have applied the pattern the deposit and import/export subsystem of the banking domain. According to our experience, we can make the same business conceptual models between the deposit and import/export subsystem due to the main concepts suggested by the pattern and effectively identify a variable business components in the UML Components development process.

• Journal of computational fluids engineering
• /
• v.11 no.3 s.34
• /
• pp.1-7
• /
• 2006

Euler and Navier-Stokes flow analyses for helicopter rotor in hover were performed as low and high fidelity analysis models respectively for the future multidisciplinary design optimization(MDO). These design-oriented analyses possess several attributes such as variable complexity, sensitivity-computation capability and modularity which analysis models involved in MDO are recommended to provide with. To realize PC-based analyses for both fidelity models, reduction of flow domain was made by appling farfield boundary condition based on 3-dimensional point sink with simple momentum theory and also periodic boundary condition in the azimuthal direction. Correlations of thrust, torque and their sensitivities between low and high complexity models were tried to evaluate the applicability of these analysis models in MDO process. It was found that the low-fidelity Euler analysis model predicted inaccurate sensitivity derivatives at relatively high angle of attack.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37 no.6A
• /
• pp.473-479
• /
• 2012

In this paper, blind equalization technique using Gaussian two-cluster model is proposed. The proposed approach, by modeling the received M-QAM signals as Gaussian distributed two-cluster, minimizes the computational complexity and enhances the reliability of the signal estimates. In addition, by using a nonlinear estimator with variable parameters to estimate the transmitted signal, and by selectively applying the reduced constellation and the original constellation when estimating the signals, the reliability of the signal estimation was further improved. As a result, the proposed approach has improved the performance while reducing the complexity of the equalizer. Through computer simulations for blind equalization of higher-order signals of 64-QAM, it was confirmed that the proposed method showed better performance than traditional approaches.