• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2001년도 가을 학술발표회 논문집
• /
• pp.515-522
• /
• 2001

A simplified method is presented for the computation of eigenvalue and eigenvector derivatives associated with repeated eigenvalues. In the proposed method, adjacent eigenvectors and orthonormal conditions are used to compose an algebraic equation whose order is (n+m)x(n+m), where n is the number of coordinates and m is the number of multiplicity of the repeated eigenvalue. One algebraic equation developed can be computed eigenvalue and eigenvector derivatives simultaneously. Since the coefficient matrix of the proposed equation is symmetric and based on N-space, this method is very efficient compared to previous methods. Moreover the numerical stability of the method is guaranteed because the coefficient matrix of the proposed equation is non-singular, This method can be consistently applied to both structural systems with structural design parameters and mechanical systems with lumped design parameters. To verify the effectiveness of the proposed method, the finite element model of the cantilever beam and a 5-DOF mechanical system in the case of a non-proportionally damped system are considered as numerical examples. The design parameter of the cantilever beam is its width, and that of the 5-DOF mechanical system is a spring.

• 한국정밀공학회지
• /
• 제17권5호
• /
• pp.180-185
• /
• 2000

An interface V-notched crack problem can be formulated as a eigenvalue problem. there are the eigenvalues which give stress singularities at the V-notched crack tip. The RWCIM is a method of calculating the eigenvector coefficients associated with eigenvalues for a V-notched crack problem. Obtaining the stress intensity factors for an interface crack in dissimilar materials is examined by the RWCIM. The results of stress intensity factors for an interface crack are compared with those of the displacement extrapolation method by the BEM

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2000년도 제15차 학술회의논문집
• /
• pp.10-10
• /
• 2000

Eigenstructure assignment is a typical method with the capability of the consideration of the specifications in time-domain in designing a 1]near control system. In general eigenstructure assignment such that all the desired eigenvalues are exactly assigned to the closed-loop system, the assignment of the eigenvectors is very restrictive. However if the arbitrary point in a certain area as an eigenvalue is allowed to be assigned t the closed-loop system, the assignment of the eigenvector corresponding to this eigenvalue can be much less restrictive. In this paper, the flexible eigenstructure assignment that can assign more closely the desired eigenvector to the closed-loop system by using an optimization technique is proposed.

• 대한수학회논문집
• /
• 제32권1호
• /
• pp.215-224
• /
• 2017

In this paper, we propose a method of order two for the computation of positive solutions to a boundary value problem of the linear beam equation. The method is based on the Power method for the eigenvector associated with the dominant eigenvalue and the Crout-like factorization algorithm for the banded system of linear equations. It is extremely fast due to the linear complexity of the linear system solver. Numerical result of a test problem is included.

• 한국토양비료학회지
• /
• 제41권6호
• /
• pp.387-392
• /
• 2008

토양의 중요한 5가지 특성 즉, 지형, 배수등급, 유효토심, 토성(속), 자갈함량 등의 특성을 이용하여 이 5가지의 특성을 복합적으로 인지하는 것은 매우 어려운 일이기 때문에 이들을 새로운 차원에 적용시켜 그 연관성들을 보고자 다차원척도법(Multidimensional Scaling)을 이용하였다. 퇴적암 유래토양의 각 특성별 구분기준은 우리나라 전체 토양에 대하여 단순화를 시도하였다. 즉, 지형요인은 15 $\rightarrow$ 9개로, 토성(속)은 32 $\rightarrow$ 6개로, 배수등급은 6 $\rightarrow$ 5개로 축합하였고, 유효토심(4개) 및 석력함량(3개)은 그대로 분석에 이용하였다. 지형요인은 산악지, 구릉지, 산록경사지, 곡간지, 선상지의 5등급으로 분류되었으며, 토성(속)은 사양질, 식양질, 미사식양질,식질의 4등급으로 분류되었으며, 배수등급은 "매우양호", "양호", "약간양호", "약간불량", "불량"의 5등급으로 분류되었다. 유효토심은 0~20, 20~50, 50~100,100cm 이상의 4등급으로 분류되었고, 자갈함량은 0~10, 10~35, 35% 이상의 3등급으로 분류되었다. 전체적으로 eigenvector면에서 볼 때 2, 3사분면에서 1, 4사분면으로 갈수록(왼쪽에서 오른쪽으로) 산토양 쪽으로 가면서 토성도 세립화하는 것을 알 수 있었다. Eigenvalue측면에서 볼 때는 뚜렷한 경향을 찾아보기는 어려웠으나 eigenvector를 기준하고 그 내에서eigenvalue를 고려하면 그룹들간의 차이가 많이 나타나는 것으로 보였다. 이와 같이 약간의 예외적인 토양이 분포함에 주시하여 분석인자의 가중치를 조정하거나, 인자의 가감등에 의하여 군집내의 동질성을 개선할 여지가 있었다. 또한 동일군집에 포함되는 토양들의 특성을 보다 면밀히 조사.연구하여 기존의 분류단위를 재고해야 할 필요성도 인정되었다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2002년도 ICCAS
• /
• pp.44.3-44
• /
• 2002

Stability robustness of a linear time-varying system with time-varying structured state space uncertainties is considered by using extended-mean theorem and Bellman's lemma. The extended-mean theorem is a necessary and sufficient exponential stability criterion based on the recently developed PD-eigenvalue and PD-eigenvector for a linear time-varying system. Our new result required that the extended-mean of each nominal PD-eigenvalue should be negative real which is determined by a norm involving the structures of the uncertainty and the no...

• 대한전기학회논문지:전력기술부문A
• /
• 제52권9호
• /
• pp.487-492
• /
• 2003

This paper analyzes an eigenvalue distribution and enhancement of the small signal stabiligy when an Unified Power Flow Controller (UPFC) modeling is connected into the multi-machine power system. Recently a lot of attention has been paid to the subject of dynamic stability. It deals with analysis of eigenvalue sensitivities with respect to parameters of UPFC Controller and damping of interarea and local electromechanical oscillation modes using UPFC Controller. It provides an insight and understanding in the basic characteristics of damping effects of UPFC Controller and shows a very stable frequency response via UPFC in test model. The series branch of the UPFC is designed to damp the power oscillation during transients, while the shunt branch aims at maintaining the bus voltage and angle. Comprehensive time-domain simulation studies using PSS/E show that the proposed robost UPFC controller can enhance the small signal stability efficiently in spite of the variations of power system operating conditions.

• 제어로봇시스템학회논문지
• /
• 제16권6호
• /
• pp.517-523
• /
• 2010

In this paper, we proposed an algorithm to detect aggressive driving status by analysing six kinds of driving patterns, which was achieved by comparing for the feature vectors using mahalanobis distance. The first step is to construct feature matrix of $6{\times}2$ size using frequency response of the time-series accelerometer data. Singular value decomposition makes it possible to find the dominant eigenvalue and its corresponding eigenvector. We use the eigenvector as the feature vector of the driving pattern. We conducted real experiments using three drivers to see the effects of recognition. Although there exists differences from individual drivers, we showed that driving patterns can be recognized with about 80% accuracy. Further research topics will include the development of aggressive driving warning system by improving the proposed technique and combining with post-processing of accelerometer signals.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2001년도 춘계학술대회 논문집 전력기술부문
• /
• pp.17-19
• /
• 2001

In this paper, eigenvalue perturbation theory and its applications for the augmented system matrix are described. This theory is quite useful in the cases where any change in a system parameter results in signifiant changes to most of the elements of the augmented matrix or where the forming of sensitivity matrix so complicate. And AMEP(augmented matrix eigenvalue perturbation) for the excitation system parameters are computed for analysis of small signal stability of KEPCO 215-machine 791-bus system.

• 한국지진공학회:학술대회논문집
• /
• 한국지진공학회 2001년도 추계 학술발표회 논문집 Proceedings of EESK Conference-Fall 2001
• /
• pp.117-124
• /
• 2001

A simplified method fur the eigenpair sensitivities of damped system with multiple eigenvalues is presented. This approach employs a reduced equation to determine the sensitivities of eigenpairs of the damped vibratory systems with multiple natural frequencies. In the proposed method, adjacent eigenvectors and orthonormal conditions are used to compute an algebraic equation whose order is (n+m)x(n+m), where n is the number of coordinates and m the number of multiplicity of multiple natural frequencies. The proposed method is an improved Lee and Jung's method which was developed previously. Two equations are used to find eigenvalue derivatives and eigenvector derivatives in Lee and Jung's method. A significant advantage of this approach over Lee and Jung's method is that one algebraic equation newly developed is enough to compute such eigenvalue derivatives and eigenvector derivatives. This method can be consistently applied to both structural systems with structural design parameters and mechanical systems with lumped design parameters. To demonstrate the theory of the proposed method and its possibilities in the case of multiple eigenvalues, the finite element model of the cantilever beam and 5-DOF mechanical system in the case of a non-proportionally damped system are considered as numerical examples. The design parameter of the cantilever beam is its height. and that of the 5-DOF mechanical system is a spring.