• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제1권1호
• /
• pp.35-43
• /
• 2001

In this paper, we propose a design method for GBSB (generalized brain-state-in-a-box) based associative memories. Based on the theoretical investigation about the properties of GBSB, we parameterize the solution space utilizing the limited number of parameters sufficient to represent the solution space and appropriate to be searched. Next we formulate the problem of finding a GBSB that can store the given pattern as stable states in the form of constrained optimization problems. Finally, we transform the constrained optimization problem into a SDP(semidefinite program), which can be solved by recently developed interior point methods. The applicability of the proposed method is illustrated via design examples.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1992년도 한국자동제어학술회의논문집(국내학술편); KOEX, Seoul; 19-21 Oct. 1992
• /
• pp.754-759
• /
• 1992

In this paper, we proposed a mixed H$_{2}$/H$_{\infty}$ optimal controller design method using the parameterization of H$_{2}$ suboptimal controller. The method is based on the minimization of H$_{2}$ performance measure with an H$_{\infty}$-norm constraint. We also derived the necessary and sufficient conditions for existence of solution from the decoupled Riccati equations. And the designed controller has state-space representation.n.

• 한국지능시스템학회논문지
• /
• 제11권7호
• /
• pp.680-688
• /
• 2001

In this paper, we propose a reliable method for searching the optimally performing generalized brain-state-in-a-box (GBSB) neural associative memory using an evolution program (EP) given a set of prototype patterns to be stored as stable equilibrium points. First, we exploit some qualitative guidelines necessary to synthesize the GBSB model. Next, we parameterize the solution space utilizing the limited number of parameters to represent the solution space. Then, we recast the synthesis of GBSB neural associative memories as two constrained optimization problems, which are equivalent to finding a solution to the original synthesis problem. Finally, we employ an evolution program (EP), which enables us to find an optimal set of parameters related to the size of domains of attraction (DOA) for prototype patterns. The validity of this approach is illustrated by a design example and computer simulations.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1996년도 Proceedings of the Korea Automatic Control Conference, 11th (KACC); Pohang, Korea; 24-26 Oct. 1996
• /
• pp.56-59
• /
• 1996

Worst-case state estimation will be proposed in this paper. By using the worst-case disturbance and worst-case state estimation, we can obtain right/left constrained coprime factors. If constrained coprime factors are used in designing a controller, the infinity-norm of closed-loop transfer matrix can be smaller than any constant .gamma.(> .gamma.$_{opt}$) without matrix dilation optimization. The derivation of left/right constrained coprime factors is achieved by doubly coprime factorization for the plant constrained by the infinity norm. And the parameterization of stabilizing controllers gives us easily understanding for H$_{\infty}$ control theory.ry.

• 제어로봇시스템학회논문지
• /
• 제11권11호
• /
• pp.907-912
• /
• 2005

In this paper, we consider the problem of designing sliding surfaces fur a class of dynamic systems with mismatched uncertainties in the state space model. In terms of LMIs, we give necessary and sufficient conditions fir the existence of a linear sliding surface such that the reduced order sliding mode dynamics is asymptotically stable and completely independent of uncertainties. We parameterize all such linear sliding surfaces by using the solution to the given LMI conditions. And, we consider the problem of designing linear sliding surfaces guaranteeing pole placement constraints or $H_2/H_infty$ performances. Finally, we give a design example in order to show the effectiveness of our method.

• Journal of Electrical Engineering and information Science
• /
• 제3권3호
• /
• pp.306-315
• /
• 1998

In this paper, a unified approach to the H\ulcorner controller design is proposed under the $\delta$-form for both continuous and discrete systems. Most of important basic concepts of H\ulcorner control, such as inner, co-inner, GCARE and GFARE, are reformulated by the unified form. The NLCF(Normalized left Comprime Factor) plant description has been reviewed in the $\delta$-form, and some corresponding results are proposed. And the unified H\ulcorner controller is designed which is based on the McFarlane and Glover{1]. The state-space parameterization for all suboptimal controllers is given under the NLCF model which may not be strictly proper, and the central controller is derived by using the solution to Hankel norm approximation problem[2]. The unified controller is applied to the industrial boiler control problem to exemplify the performance of the controller.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2007년도 정기 학술대회 논문집
• /
• pp.577-582
• /
• 2007

A variational formulation for plane elasticity problems is derived based on an isogeometric approach. The isogeometric analysis is an emerging methodology such that the basis functions in analysis domain arc generated directly from NURBS (Non-Uniform Rational B-Splines) geometry. Thus. the solution space can be represented in terms of the same functions to represent the geometry. The coefficients of basis functions or the control variables play the role of degrees-of-freedom. Furthermore, due to h-. p-, and k-refinement schemes, the high order geometric features can be described exactly and easily without tedious re-meshing process. The isogeometric sensitivity analysis method enables us to analyze arbitrarily shaped structures without re-meshing. Also, it provides a precise construction method of finite element model to exactly represent geometry using B-spline base functions in CAD geometric modeling. To obtain precise shape sensitivity, the normal and curvature of boundary should be taken into account in the shape sensitivity expressions. However, in conventional finite element methods, the normal information is inaccurate and the curvature is generally missing due to the use of linear interpolation functions. A continuum-based adjoint sensitivity analysis method using the isogeometric approach is derived for the plane elasticity problems. The conventional shape optimization using the finite element method has some difficulties in the parameterization of boundary. In isogeometric analysis, however, the geometric properties arc already embedded in the B-spline shape functions and control points. The perturbation of control points in isogeometric analysis automatically results in shape changes. Using the conventional finite clement method, the inter-element continuity of the design space is not guaranteed so that the normal vector and curvature arc not accurate enough. On tile other hand, in isogeometric analysis, these values arc continuous over the whole design space so that accurate shape sensitivity can be obtained. Through numerical examples, the developed isogeometric sensitivity analysis method is verified to show excellent agreement with finite difference sensitivity.

• 한국전산구조공학회논문집
• /
• 제20권3호
• /
• pp.339-345
• /
• 2007

본 논문에서는 등기하 해석법을 이용하여 평면 탄성문제의 변분식을 유도하였다. 등기하 해석법은 새로이 부각되고 있는 해석법으로서 기저 함수가 NURBS(Non-Uniform Rational B-Splines) 로부터 직접 생성되므로 해 공간은 CAD 모델을 구성하는 함수로써 표현된다. 또한 CAD 모델의 B-Spline 기저 함수를 직접 사용하므로 기하학적으로 엄밀한 형상을 표현할 수 있고 요소망의 재구성 없이 해석모델을 정밀화(Refinement)할 수 있는 강점이 있다. 본 논문에서는 이를 확장하여 연속체 기반의 애드조인트 설계 민감도 해석법을 사용하는 등기하 설계민감도 해석법을 유도하였다. 기존의 유한요소 기반형상 최적설계는 형상의 매개화에 어려움을 겪었으나 등기하 기반 최적설계에서는 기하학적 정보가 이미 B-spline 기저함수와 조정점에 포함되어 있으므로 이러한 어려움을 피할 수 있는 잠재력을 가지고 있다. 몇몇 수치 예제를 통해서 등기하 해석법을 사용한 설계 민감도 해석을 수행하였으며 유한차분 민감도와 비교하여 정확성을 확인하였다.

• 한국전산구조공학회논문집
• /
• 제24권1호
• /
• pp.1-7
• /
• 2011

본 논문에서는 아이소-지오메트릭 해석법을 이용하여 설계 의존형 하중조건을 갖는 구조물에 대한 형상 최적설계를 수행하였다. 유한요소법 기반 형상 최적설계는 CAD와 해석 모델의 차이로 인해, 설계영역 매개 변수화에 어려움이 있다. 아이소-지오메트릭 해석법은 CAD 모델과 동일한 NURBS 기저 함수와 조정점을 해석에 이용함으로써 설계의 기하학적 변화를 해석모델에 직접적으로 표현할 수 있는 장점을 가진다. 하중조건이 설계 영역에 따라 변화하는 최적설계 문제의 경우, 정확한 설계 영역 표현은 법선 벡터, 즉 변화하는 하중의 방향, 곡률 등 고차항의 정보를 정확하게 표현할 수 있고, 따라서 목적함수를 최소 또는 최대화시키는 최적의 해로 이끌어 낸다. 유한요소법 또는 밀도법을 이용한 형상 최적설계에서 설계 의존형 하중조건을 갖는 구조물의 문제를 푸는 경우, 최적설계가 진행됨에 있어 변화하는 경계의 부정확성 때문에 정확한 설계민감도를 얻기가 어려운 점이 있다. 본 논문에서는 수치 예제를 통해 아이소-지오메트릭 설계민감도를 활용한 형상 최적설계 기법이 설계 의존형 하중조건을 갖는 구조물 문제에서 유한요소 기반의 최적설계보다 더 나은 결과를 제시함을 확인하였다.