• 제목/요약/키워드: linear convergence

검색결과 1,321건 처리시간 0.026초

Simplified neuron functions for FPGA evaluations of engineering neuron on gate array and analogue circuit

  • Saito, Masayuki;Wang, Qianyi;Aoyama, Tomoo
    • 제어로봇시스템학회:학술대회논문집
    • /
    • 제어로봇시스템학회 2001년도 ICCAS
    • /
    • pp.157.6-157
    • /
    • 2001
  • We estimated various neuron functions to construct of engineering neurons, which are the combination of sigmoid, linear, sine, quadric, double/single bended, soft max/minimum functions. These combinations are estimated by the property on the potential surface between the learning points, calculation speed, and learning convergence; because the surface depends on the inference ability of a neuron system; and speed and convergence are depend on the efficiency on the points of engineering applications. After the evaluating discussions, we can select more appropriate combination than original sigmoid function´s, which is single bended function and linear one. The combination ...

  • PDF

POLYNOMIAL CONVERGENCE OF PRIMAL-DUAL ALGORITHMS FOR SDLCP BASED ON THE M-Z FAMILY OF DIRECTIONS

  • Chen, Feixiang
    • Journal of applied mathematics & informatics
    • /
    • 제30권1_2호
    • /
    • pp.127-133
    • /
    • 2012
  • We establish the polynomial convergence of a new class of path-following methods for SDLCP whose search directions belong to the class of directions introduced by Monteiro [3]. We show that the polynomial iteration-complexity bounds of the well known algorithms for linear programming, namely the short-step path-following algorithm of Kojima et al. and Monteiro and Alder, carry over to the context of SDLCP.

선형시스템을 위한 개선된 수렴속도를 갖는 기준모델 적응제어 (Model Reference Adaptive Control for Linear System with Improved Convergence Rate-parameter Adaptation Method)

  • Lim, Kye-Young
    • 대한전기학회논문지
    • /
    • 제37권12호
    • /
    • pp.884-893
    • /
    • 1988
  • Adaptive controllers for linear unknown coefficient system, that is corrupted by disturbance, are designed by parameter adaptation model reference adaptive control(MRAC). This design is stemmed from the Lyapunov direct method. To reduce the model following error and to improve the convergence rate of the design, an indirect-suboptimal control law is derived. Proper compensation for the effects of time-varying coefficients and plant disturbance are suggested. In the design procedure no complete identification of unknown coefficients are required.

  • PDF

LMI 가능성 문제를 위한 타원 알고리즘의 개선 (An improved ellipsoid algorithm for LMI feasibility problems)

  • 방대인;최진영
    • 대한전기학회:학술대회논문집
    • /
    • 대한전기학회 2002년도 합동 추계학술대회 논문집 정보 및 제어부문
    • /
    • pp.188-192
    • /
    • 2002
  • The ellipsoid algorithm solves some feasibility(or optimization) problems with LMI(Linear Matrix Inequality) constraint in polynomial time. Recently, it has been replaced by interior point algorithm due to its slow convergence and incapability of verifying feasibility. This paper proposes a method to improve its convergence by using the deep-cut method of linear programming. Simulation results show that the improved algorithm is more effective than the original one.

  • PDF

한계상태를 고려한 구조물의 최적화에 관한 연구 (A Study on Optimization of Structure with Limit State Constraints)

  • 김기대
    • 한국산업융합학회 논문집
    • /
    • 제7권2호
    • /
    • pp.181-186
    • /
    • 2004
  • This study presents a optimization of structure, in which constraints contain the conditions of stress and serviceability, while the sequential linear programming method (SLP) is used as a rational approach. The optimum design results contained on the limit state constraints are compared with those obtained by the only stress and ministry of construction enacted standard plans. A simple slab bridge is analysed numerically for illustration of the structural optimization. It may be asserted that serviceability constraints is very important to a structure design.

  • PDF

PRECONDITIONED AOR ITERATIVE METHOD FOR Z-MATRICES

  • Wang, Guangbin;Zhang, Ning;Tan, Fuping
    • Journal of applied mathematics & informatics
    • /
    • 제28권5_6호
    • /
    • pp.1409-1418
    • /
    • 2010
  • In this paper, we present a preconditioned iterative method for solving linear systems Ax = b, where A is a Z-matrix. We give some comparison theorems to show that the rate of convergence of the new preconditioned iterative method is faster than the rate of convergence of the previous preconditioned iterative method. Finally, we give one numerical example to show that our results are true.

NONPARAMETRIC DISCONTINUITY POINT ESTIMATION IN GENERALIZED LINEAR MODEL

  • Huh, Jib
    • Journal of the Korean Statistical Society
    • /
    • 제33권1호
    • /
    • pp.59-78
    • /
    • 2004
  • A regression function in generalized linear model may have a discontinuity/change point at unknown location. In order to estimate the location of the discontinuity point and its jump size, the strategy is to use a nonparametric approach based on one-sided kernel weighted local-likelihood functions. Weak convergences of the proposed estimators are established. The finite-sample performances of the proposed estimators with practical aspects are illustrated by simulated examples.

근사 복원 IIR/FIR QMF 필터뱅크에서 지연요소 변화에 따른 특성에 관한 연구 (A Study on the Characteristics of Delay Parameter Change in Approximately Reconstruction IIR/FIR QMF Filter Banks)

  • 이상준;김남수;김남호
    • 융합신호처리학회 학술대회논문집
    • /
    • 한국신호처리시스템학회 2003년도 하계학술대회 논문집
    • /
    • pp.296-299
    • /
    • 2003
  • This paper shows a novel and simple IIR/FIR QMF(quadrature mirror filter) filter banks, mixed IIR and FIR structure. Here, FIR filters used for phase compensation. In this paper, we introduced analysis and synthesis filter banks, which used FIR linear phase filters and all pass filters. In result, phase response of analysis and synthesis filter banks become approximately linear characteristic. Simultaneously, a liasing distortion can be completely canceled.

  • PDF

피드백시스템에 대한 직접학습제어의 응용 (Application of Direct Learning Control to Feedback Systems)

  • 안현식
    • 융합신호처리학회 학술대회논문집
    • /
    • 한국신호처리시스템학회 2003년도 하계학술대회 논문집
    • /
    • pp.173-176
    • /
    • 2003
  • In this paper, a DLC method is suggested for linear feedback systems to improve the tracking performance when the task of the system is repetitive. DLC can generate the desired control input directly from the previously teamed control inputs corresponding to other output trajectories. It is assumed that all outputs considered in this paper have some relations called "proportionality. " To show the validity and tracking performance of the proposed method, some simulation are performed for the tracking control of a linear system with a PI controller.

  • PDF