• 상하수도학회지
• /
• 제12권3호
• /
• pp.99-106
• /
• 1998

This paper presents a procedure for determining the design rainfall depth and the design rainfall intensity at Incheon city area in Korea. In this study the eight probability distributions are considered to estimate the probable rainfall depths for 11 different durations. The Kolmogorov - Smirnov test and the Chi-square test are adopted to test each distribution. The probable rainfall intensity formulas are then determined by i) the least squares (LS) method, ii) the least median squares (LMS) method, iii) the reweighted least squares method based on the LMS (RLS), and iv) the constrained regression (CR) model. The Talbot, the Sherman, the Japanese, and the Unified type are considered to determine the best type for the Incheon station. The root mean squared (RMS) errors are computed to test the formulas derived by four methods. It is found that the Unified type is the most reliable and that all methods presented herein are acceptable for determining the coefficients of rainfall intensity formulas from an engineering point of view.

• 제어로봇시스템학회논문지
• /
• 제9권2호
• /
• pp.159-167
• /
• 2003

This study suggests the hybrid fault diagnosis method of signed digraph (SDG) and partial least squares (PLS). SDG offers a simple and graphical representation for the causal relationships between process variables. The proposed method is based on SDG to utilize the advantage that the model building needs less information than other methods and can be performed automatically. PLS model is built on local cause-effect relationships of each variable in SDG. In addition to the current values of cause variables, the past values of cause and effect variables are inputted to PLS model to represent the Process armies. The measured value and predicted one by dynamic PLS are compared to diagnose the fault. The diagnosis example of CSTR shows the proposed method improves diagnosis resolution and facilitates diagnosis of masked multiple-fault.

• 전력전자학회논문지
• /
• 제23권1호
• /
• pp.72-75
• /
• 2018

An initial rotor position estimation method is proposed in this study for an interior permanent-magnet synchronous motor without a resolver or an absolute encoder. This method uses least squares approximation to estimate the initial rotor position. The magnetic polarity is identified by injection of short pulses. The proposed estimation process is robust because it does not require complex signal processing that depends on the performance of a digital filter. In addition, it can be applied to various servo systems because it does not require additional hardware. Experimental results validate the effectiveness of the proposed method using a standard industrial servomotor with interior-permanent magnets.

• Communications for Statistical Applications and Methods
• /
• 제2권2호
• /
• pp.135-145
• /
• 1995

A new algorithm was given to successively fit the multiexponential function/nonlinear function to data by a weighted least squares method, using Gauss-Newton, Marquardt, gradient and DUD methods for convergence. This study also considers the problem of linear-nonlimear weighted least squares estimation which is based upon the usual Taylor's formula process.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1992년도 하계학술대회 논문집 A
• /
• pp.319-322
• /
• 1992

This paper proposes indirect adaptive pole placement adaptive controller using a modified least squares method. If an adaptive controller has good performance, it is necessary that an estimator have fast convergence. This paper presents a modified least squares method which guarantees the stability of estimator and has fast convergence. In this algorithm, information on signal level is obtained from the determinent of covariance matrix and according to it, weighting factor is tuned.

• 충청수학회지
• /
• 제22권2호
• /
• pp.177-186
• /
• 2009

The global least squares method (Gl-LSQR) is a generalization of LSQR method for solving linear system with multiple right hand sides. In this paper, we present how to apply this algorithm for solving the image restoration problem and illustrate the usefulness and effectiveness of this method from numerical experiments.

• 한국측량학회지
• /
• 제36권6호
• /
• pp.621-628
• /
• 2018

GPS (Global Positioning System)를 이용하여 위치를 결정하기 위해서는 4개 이상의 가시위성이 있어야 한다. 하지만 도심지역과 같은 환경에서는 이러한 조건을 만족하기 어려운 경우도 있다. 특히, 가시위성이 3개뿐인 경우 외부로부터 위치결정에 필요한 시계오차정보를 활용하는 측위기법이 대안으로 사용되기도 한다. 본 연구에서는 먼저 수신기 시계오차특성을 분석한 후 시계오차의 보간에 적합한 방법으로 LSC (Least-Squares Collocation)을 제안하였다. 실험을 위해 국내 상시관측소와 상시관측소 근처에 설치된 수신기로부터 수신된 GPS 데이터를 이용하였다. DGPS (Differential GPS)기법을 통해 먼저 시계오차를 계산했으며 효율적인 보간을 위해 구간을 나눈 후 보간하는 방법을 적용하였다. 시계오차의 계산이 불가능한 epoch에 대해 LSC 보간법을 적용함으로써 시계오차를 계산하였다. 실험결과를 분석하기 위해 원래 데이터로부터 계산된 시계오차와 보간된 시계오차와의 차이인 잔차를 계산하였다. 계산결과 잔차의 평균은 0.24m 그리고 표준편차는 0.49m로 충분한 정확도의 확보가 가능한 것으로 판단된다.

• 전자공학회논문지
• /
• 제50권12호
• /
• pp.150-158
• /
• 2013

본 논문에서는 절단된 제약적 최소 제곱 필터를 이용하여 방향 적응적으로 영상을 복원하는 방법을 제안한다. 제안하는 영상복원 필터는 공간영역에서 이론적으로 영상 전체의 크기를 갖는 제약적 최소 제곱(constrained least squares; CLS) 필터를 Maxwell-Boltzmann 커널을 사용하여 절단한 유한 임펄스 응답(finite impulse response; FIR) 필터의 구조로 실시간 영상복원을 가능하게 한다. 또한 화소 단위로 공분산 행렬을 분석하여 방향성을 추정하여, 화소마다 필터의 계수를 적응적으로 생성하여 방향 적응적으로 영상을 복원한다. 실험결과를 통해 기존의 알고리듬에 비해 제안된 방법이 선명하고 부작용(artifacts)이 적은 결과를 얻는 것을 검증하였다.

• 지구물리와물리탐사
• /
• 제8권2호
• /
• pp.170-176
• /
• 2005

속도중합의 역산을 이용하면 탄성파 자료처리에 있어서 다양한 처리가 가능하므로 이 분야는 최근에 들어 매우 유용한 영역으로 주목을 받고 있다. 하지만 다양한 처리에 적용하기 위해서는 사용되는 역산 방법이 잡음에 강하면서도 고해상도의 속도중합 결과를 만들 수 있어야 한다. 이러한 특성을 갖는 대표적인 역산에는 ${L_1}-norm$을 최소화시키는 IRLS(Iteratively Reweighted Least-Squares)방법을 주로 사용하였다. 본 논문에서는 이러한 성질을 갖는 또 다른 역산 방법의 하나로서 CGG (Conjugate Guided Gradient) 방법을 소개한다. CGG 방법은 반복적 최소자승법의 하나인 Conjugate Gradient (CG)방법을 변형시킨 형태로 ${L_1}-norm$을 최소화 시키는 역산법으로 활용할 수 있다. 본 논문에서는 CGG방법을 소개하고 기존의 IRLS방법과의 차이점 및 결과들을 비교하였다. 모의자료와 현장자료에 대한 실험결과를 통해서 CGG 방법이 IRLS방법과 마찬가지로 다양한 잔여/모델 norm을 최소화시키는 역산방법으로 사용될 수 있음을 보여준다.

• 센서학회지
• /
• 제26권6호
• /
• pp.371-378
• /
• 2017

An accurate and robust estimation of center of rotation (CoR) using optical marker trajectory is crucial in human biomechanics. In this regard, the performances of the two prevailing least-squares methods, the Gamage and Lasenby (GL) method, and the Chang and Pollard (CP) method, are verified in this paper. While both methods are sphere-fitting approaches in closed form and require no tuning parameters, they have not been thoroughly verified by comparison of their estimation accuracies. Furthermore, while for both methods, results for stationary CoR locations are presented, cases for perturbed CoR locations have not been investigated for any of them. In this paper, the estimation performances of the GL method and CP method are investigated by varying the range of motion (RoM) and noise amount, for both stationary and perturbed CoR locations. The difference in the estimation performance according to the variation in the amount of noise and RoM was clearly shown for both methods. However, the CP method outperformed the GL method, as seen in results from both the simulated and the experimental data. Particularly, when the RoM is small, the GL method failed to estimate the appropriate CoR while the CP method reasonably maintained the accuracy. In addition, the CP method showed a considerably better predictability in CoR estimation for the perturbed CoR location data than the GL method. Accordingly, it may be concluded that the CP method is more suitable than the GL method for CoR estimation when RoM is limited and CoR location is perturbed.