• Journal of information and communication convergence engineering
• /
• 제8권3호
• /
• pp.267-272
• /
• 2010

For the last decade, recurrent neural networks (RNNs) have been commonly applied to communications channel equalization. The major problems of gradient-based learning techniques, employed to train recurrent neural networks are slow convergence rates and long training sequences. In high-speed communications system, short training symbols and fast convergence speed are essentially required. In this paper, the derivative-free Kalman filter, so called the unscented Kalman filter (UKF), for training a fully connected RNN is presented in a state-space formulation of the system. The main features of the proposed recurrent neural equalizer are fast convergence speed and good performance using relatively short training symbols without the derivative computation. Through experiments of nonlinear channel equalization, the performance of the RNN with a derivative-free Kalman filter is evaluated.

• 한국터널지하공간학회 논문집
• /
• 제19권5호
• /
• pp.799-812
• /
• 2017

쉴드 TBM 터널에 대한 2차원 수치해석 시 기존 연구에서는 뒤채움재 및 실제적인 내공변위 분포를 적절히 고려하지 않았다. 따라서 본 연구에서는 이를 보완하고자 gap parameter 적용 시 굴착면에서 발생하는 실제적인 내공변위분포를 고려하고, 적정한 뒤채움재 물성치를 적용한 모델링 방법을 제시하였다. 이를 위해 단일층 풍화토 지반을 대상으로 gap parameter와 토피고가 굴착면의 지반손실량과 지표손실량으로 구성되는 체적손실량에 미치는 영향을 분석하였다. 그 결과 대부분의 지표침하가 굴착 직후에 발생하였으며, 토피고 및 gap parameter가 증가할수록 수치해석을 통해 얻은 지표침하곡선의 영향범위가 제안된 이론식보다 넓게 나타났다. 따라서 2차원 수치해석 시 쉴드 TBM 터널의 거동을 정확히 평가하기 위해서는 하중분담률을 이용하여 gap parameter를 적용하고, 뒤채움재의 물성치도 적절히 고려해야 한다고 판단된다.

• 전기학회논문지
• /
• 제58권9호
• /
• pp.1769-1774
• /
• 2009

An algorithm for extracting the BJT DC model parameter values for SPICE model is proposed. The nonlinear optimization method for analyzing the device I-V data using the Levenberg-Marquardt algorithm is proposed and the method for calculating initial conditions of model parameters to improve the convergence characteristics is proposed. The base current and collector current obtained from the proposed method shows the root mean square error of 6.04% compared with the measured data of the PNP BJT named 2SA1980.

• 한국산업융합학회 논문집
• /
• 제25권5호
• /
• pp.853-859
• /
• 2022

In this study, we analyzed the instantaneous trajectory generation capability and target arrival rate of a mobile robot by changing the parameter of the TEB (Timed Elastic Band) Local Planner among local planners that affect the instantaneous obstacle avoidance ability of the mobile robot using ROS (Robot Operating System) simulation and real experience. As a result, we can expect a decrease in the target arrival time of the mobile robot through a decrease in the parameter values of the TEB Local Planner's min_obstacle_dist, inflation_dist, and penalty_epsilon. However, if this parameter is reduced too much, the risk of obstacle collision of the moving robot is increases, so it is important to combine the appropriate values to construct the parameter.

• 한국통신학회논문지
• /
• 제24권9A호
• /
• pp.1329-1338
• /
• 1999

본 논문에서, LMS 알고리즘의 수렴 속도를 향상시키기 위한 효율적인 신호간섭 제어기법을 제안한다. 수신 데이터를 재활용하여 심볼 시간 주기에 계수들을 곱함으로써 적응되는 제안된 알고리즘의 수렴특성이 수렴 속도의 향상을 이론적으로 증명하기 위해 분석한다. 스텝-크기 매개변수 $\mu$가 증가됨에 따라 알고리즘의 수렴 속도가 제어된다. 또한, 스텝-크기 매개변수 $\mu$의 증가는 실험적으로 계산된 학습 곡선에서 분산을 감소시키는 효과를 갖는다. 고유치 확산을 증가시킴에 따라 적응 등화기의 수렴속도를 천천히 제어하고 평균 자승 에러의 안정-상태 값을 증가시키는 효과를 나타내며 데이터-재사용 LMS 기술이 수렴속도를 (B+1)배만큼 증가시켜 필터 알고리즘에서 신호간섭제어의 우수성을 입증한다.

• ETRI Journal
• /
• 제35권6호
• /
• pp.1011-1020
• /
• 2013

In this paper, we present a method for obtaining a high-quality 3D depth. The advantages of active pattern projection and passive stereo matching are combined and a system is established. A diffractive optical element (DOE) is developed to project the active pattern. Cross guidance (CG) and auto guidance (AG) are proposed to perform the passive stereo matching in a stereo image in which a DOE pattern is projected. When obtaining the image, the CG emits a DOE pattern periodically and consecutively receives the original and pattern images. In addition, stereo matching is performed using these images. The AG projects the DOE pattern continuously. It conducts cost aggregation, and the image is restored through the process of removing the pattern from the pattern image. The ground truth is generated to estimate the optimal parameter among various stereo matching algorithms. Using the ground truth, the optimal parameter is estimated and the cost computation and aggregation algorithm are selected. The depth is calculated and bad-pixel errors make up 4.45% of the non-occlusion area.

• 한국융합학회논문지
• /
• 제11권4호
• /
• pp.63-69
• /
• 2020

본 연구는 복부압박벨트가 만성 뇌졸중 환자의 안정성한계와 보행 변수에 즉각적인 효과를 알아보기 위해 실시하였다. 뇌졸중 환자 30명을 모집하여 복부압박벨트 착용 전과 복부압박벨트 착용 후 변화를 확인하였다. 측정은 균형변수인 안정성한계(limited of stability), 시공간적 보행 변수(spatiotemporal gait parameter)를 측정하였다. 복부압박벨트 착용 후 마비측, 비마비측, 전방, 후방 이동면적이 유의하게 증가하였고(p<.05), 시간적 보행 변수인 분속수(cadence), 보행속도(gait velocity), 공간적 보행변수인 보폭(stride length)이 유의하게 증가하였다(p<.05). 본 연구를 통해 복부압박벨트 착용은 뇌졸중 환자에게 균형과 보행 기능 개선에 즉각적인 효과가 있다는 것을 확인할 수 있었다. 향후 연구에서는 뇌졸중 환자의 균형과 보행 기능 개선에 효율적인 복부압박 수준과 중재기간에 대한 연구가 필요하다.

• Journal of Mechanical Science and Technology
• /
• 제15권10호
• /
• pp.1356-1368
• /
• 2001

In this paper, the model reference adaptive control (MRAC) of a flexible structure is investigated. Any mechanically flexible structure is inherently distributed parameter in nature, so that its dynamics are described by a partial, rather than ordinary, differential equation. The MRAC problem is formulated as an initial value problem of coupled partial and ordinary differential equations in weak form. The well-posedness of the initial value problem is proved. The control law is derived by using the Lyapunov redesign method on an infinite dimensional filbert space. Uniform asymptotic stability of the closed loop system is established, and asymptotic tracking, i. e., convergence of the state-error to zero, is obtained. With an additional persistence of excitation condition for the reference model, parameter-error convergence to zero is also shown. Numerical simulations are provided.

• Journal of applied mathematics & informatics
• /
• 제29권3_4호
• /
• pp.813-829
• /
• 2011

In this article, we present a numerical scheme for solving singularly perturbed (i.e. highest -order derivative term multiplied by small parameter) Burgers-Huxley equation with appropriate initial and boundary conditions. Most of the traditional methods fail to capture the effect of layer behavior when small parameter tends to zero. The presence of perturbation parameter and nonlinearity in the problem leads to severe difficulties in the solution approximation. To overcome such difficulties the present numerical scheme is constructed. In construction of the numerical scheme, the first step is the dicretization of the time variable using forward difference formula with constant step length. Then, the resulting non linear singularly perturbed semidiscrete problem is linearized using quasi-linearization process. Finally, differential quadrature method is used for space discretization. The error estimate and convergence of the numerical scheme is discussed. A set of numerical experiment is carried out in support of the developed scheme.

• 대한전기학회논문지
• /
• 제38권7호
• /
• pp.557-568
• /
• 1989

Using Power Spectrum Analysis, conditions of parameter convergence for a Model Reference Adaptive Control (MRAC) system are described. The general Persistent Excitation (PE) condition given in time domain can be transformed to the positiveness of auto-correlation matrix which is represented in frequency domain by the spectra of reference input signal. For an MRAC system designed with relative degree one, the existence and the uniqueness of parameter nominal values due to the variation of input spectra can be analyzed by the PE condition in frequency domain. If the input signal has 2n spectra or more, it can be shown that the nominal values exist independent of adaptive gain, input amplitudes, and magnitudes or numbers of their spectra.