• 제목/요약/키워드: Moving Least Square

검색결과 125건 처리시간 0.038초

A New Least Mean Square Algorithm Using a Running Average Process for Speech Enhancement

  • Lee, Soo-Jeong;Ahn, Chan-Sik;Yun, Jong-Mu;Kim, Soon-Hyob
    • The Journal of the Acoustical Society of Korea
    • /
    • 제25권3E호
    • /
    • pp.123-130
    • /
    • 2006
  • The adaptive echo canceller (AEC) has become an important component in speech communication systems, including mobile station. In these applications, the acoustic echo path has a long impulse response. We propose a running-average least mean square (RALMS) algorithm with a detection method for acoustic echo cancellation. Using colored input models, the result clearly shows that the RALMS detection algorithm has a convergence performance superior to the least mean square (LMS) detection algorithm alone. The computational complexity of the new RALMS algorithm is only slightly greater than that of the standard LMS detection algorithm but confers a major improvement in stability.

캐널곡면의 복원 (Reconstruction of Canal Surfaces)

  • 이인권;김구진
    • 한국정보과학회논문지:시스템및이론
    • /
    • 제32권8호
    • /
    • pp.411-417
    • /
    • 2005
  • 본 논문에서는 입력으로 주어진 점집차 (point cloud)으로부터 캐널곡면을 복원 (reconstruction)하는 알고리즘을 제시한다. 캐널곡면은 반경이 변화하며 중심점이 기준곡선(spine curve)을 따라 이동하는 구(moving sphere)의 스웹트곡면(swept surface)이다. 이 논문에서는 수축기법(shrinking method)과 moving least-square 방법을 이용하여 주어진 점집합을 세곡선(thin-curve)형태의 점집합으로 수축시킴으로써 캐널곡면의 기준곡선을 근사한다. 근사된 기준곡선과 입력으로 주어진 점집합에 포함된 점들 사이의 거리를 이용하여, 캐널곡면을 구성하는 이동 구의 반경을 계산한다.

데이터베이스 기반 유전 알고리즘을 이용한 효율적인 에어포일 형상 최적화에 대한 연구 (Study on the Airfoil Shape Design Optimization Using Database based Genetic Algorithms)

  • 권장혁;김진;김수환
    • Journal of Advanced Marine Engineering and Technology
    • /
    • 제31권1호
    • /
    • pp.58-66
    • /
    • 2007
  • Genetic Algorithms (GA) have some difficulties in practical applications because of too many function evaluations. To overcome these limitations, an approximated modeling method such as Response Surface Modeling(RSM) is coupled to GAs. Original RSM method predicts linear or convex problems well but it is not good for highly nonlinear problems cause of the average effect of the least square method(LSM). So the locally approximated methods. so called as moving least squares method(MLSM) have been used to reduce the error of LSM. In this study, the efficient evolutionary GAs tightly coupled with RSM with MLSM are constructed and then a 2-dimensional inviscid airfoil shape optimization is performed to show its efficiency.

실시간 공칭 모델 추정 외란관측기에 관한 실험 연구: 재귀최소자승법 (An Experimental Study on Realtime Estimation of a Nominal Model for a Disturbance Observer: Recursive Least Squares Approach)

  • 이상덕;정슬
    • 제어로봇시스템학회논문지
    • /
    • 제22권8호
    • /
    • pp.650-655
    • /
    • 2016
  • In this paper, a novel RLS-based DOB (Recursive Least Squares Disturbance Observer) scheme is proposed to improve the performance of DOB for nominal model identification. A nominal model can be generally assumed to be a second order system in the form of a proper transfer function of an ARMA (Autoregressive Moving Average) model. The RLS algorithm for the model identification is proposed in association with DOB. Experimental studies of the balancing control of a one-wheel robot are conducted to demonstrate the feasibility of the proposed method. The performances between the conventional DOB scheme and the proposed scheme are compared.

Algorithm for Detection of Fire Smoke in a Video Based on Wavelet Energy Slope Fitting

  • Zhang, Yi;Wang, Haifeng;Fan, Xin
    • Journal of Information Processing Systems
    • /
    • 제16권3호
    • /
    • pp.557-571
    • /
    • 2020
  • The existing methods for detection of fire smoke in a video easily lead to misjudgment of cloud, fog and moving distractors, such as a moving person, a moving vehicle and other non-smoke moving objects. Therefore, an algorithm for detection of fire smoke in a video based on wavelet energy slope fitting is proposed in this paper. The change in wavelet energy of the moving target foreground is used as the basis, and a time window of 40 continuous frames is set to fit the wavelet energy slope of the suspected area in every 20 frames, thus establishing a wavelet-energy-based smoke judgment criterion. The experimental data show that the algorithm described in this paper not only can detect smoke more quickly and more accurately, but also can effectively avoid the distraction of cloud, fog and moving object and prevent false alarm.

일반화 최소자승추정의 시공간경사법에 의한 실시간 자동목표 추적 (Real-Time Automatic Target Tracking Based on Spatio-Temporal Gradient Method with Generalized Least Square Estimation)

  • 장익훈;김종대;김남철;김재균
    • 대한전자공학회논문지
    • /
    • 제26권1호
    • /
    • pp.78-87
    • /
    • 1989
  • 본 논문에서는 백색 Gauss 잡음이 섞인 연속영상으로 부터 물체의 이동정보를 검출하기 위하여 최소자승추정의 시공간경사법을 제안하였다. 제안된 방법은 하나의 이동물체를 실시간으로 추적하도록 고속의 16-bit 마이크로 프로세서를 사용한 자동목표 추적장치에 적용되었다. 실험결과 제안된 방법은 기존의 최소자승추정의 시공간경사법에 비해서 매우 우수한 성능을 보였다.

  • PDF

임의의 절점 추가에 의한 개선 유한요소법 (An Improved Finite Element Method by Adding Arbitrary Nodes in a Domain)

  • 김현규
    • 대한기계학회논문집A
    • /
    • 제30권12호
    • /
    • pp.1626-1633
    • /
    • 2006
  • In the present paper, in the context of the meshless interpolation of a moving least squares (MLS) type, a novel method which uses primary and secondary nodes in the domain and on the global boundary is introduced, in order to improve the accuracy of solution. The secondary nodes can be placed at any location where one needs to obtain a better resolution. The support domains for the shape functions in the MLS approximation are defined from the primary nodes, and the secondary nodes use the same support domains. The shape functions based on the MLS approximation, in an integration domain, have a single type of a rational function, which reduces the difficulty of numerical integration to evaluate the weak form. The present method is very useful in an adaptive calculation, because the secondary nodes can be easily added and moved without an additional mesh. Several numerical examples are presented to illustrate the effectiveness of the present method.

Redistance 방정식의 경계조건이 슬로싱 문제의 level set 해석에 미치는 영향 (EFFECT OF THE BOUNDARY CONDITION OF REDISTANCE EQUATION ON THE LEVEL SET SOLUTION OF SLOSHING PROBLEM)

  • 최형권
    • 한국전산유체공학회:학술대회논문집
    • /
    • 한국전산유체공학회 2009년 춘계학술대회논문집
    • /
    • pp.165-169
    • /
    • 2009
  • The effect of the Dirichlet boundary condition for the redistance equation of level set method on the solutionof sloshing problem is investigated by adopting four Dirichlet boundary conditions. For the solution of the incompressible Navier-Stokes equations, P1P1 four-step fractional finite element method is employed and a least-square finite element method is used for the solutions of the two hyperbolic type equations of level set method; advection and redistance equation. ALE (Arbitrary Lagrangian Eulerian) method is used to deal with a moving computational domain. It has been shown that the free surface motion in a sloshing tank is strongly dependent on the type of the Dirichlet boundary condition and the results of broken dam and sloshing problems using various Dirichlet boundary conditions are discussed and compared with the existing experimental results.

  • PDF