• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송∙미디어공학회 2020년도 추계학술대회
• /
• pp.241-243
• /
• 2020

The high computational complexity of deep learning algorithms has led to the development of specialized hardware architectures. However, soft errors (bit flip) may occur in these hardware systems due to voltage variation and high energy particles. Many error correction methods have been proposed to counter this problem. In this work, we analyze an error correction mechanism based on repetition codes and an activation function. We test this method by injecting errors into weight filters and define an ideal error rate range in which the proposed method complements the accuracy of the model in the presence of error.

• 전력전자학회:학술대회논문집
• /
• 전력전자학회 2008년도 하계학술대회 논문집
• /
• pp.247-249
• /
• 2008

There are many types of grid-connected inverter controllers; Synchronous Reference Frame (SRF)-based controller is the most popular methods. SRF-based controller is capable for reducing both of zero-steady state error and phase delay. However, SRF-based controller has a complex algorithm to apply in real application such as digital processor. Resonant controller is also reduced zero-steady state error, but its transfer function has a high order. In this paper, a simple proportional control is applied for grid connected inverter with LCL filter. LCL filter is a third order system. Applying a simple proportional controller is not increased the order of closed loop transfer function. By this technique, the single phase model is easily obtained. To reduce steady state error, proportional gain is set as high as possible, but it may produce instability. To compromise between a minimum steady state error and stability, the single phase model is evaluate through Root Locus and Bode diagram. PSIM simulation is used to verify the analysis.

• ETRI Journal
• /
• 제34권5호
• /
• pp.696-705
• /
• 2012

Accurate synthesis models based on artificial neural networks (ANNs) are proposed to directly obtain the physical dimensions of an asymmetric coplanar waveguide with conductor backing and substrate overlaying (ACPWCBSO). First, the ACPWCBSO is analyzed with the conformal mapping technique (CMT) to obtain the training data. Then, a modified genetic-algorithm-Levenberg-Marquardt (GA-LM) algorithm is adopted to train ANNs. In the algorithm, the maximal relative error (MRE) is used as the fitness function of the chromosomes to guarantee that the MRE is small, while the mean square error is used as the error function in LM training to ensure that the average relative error is small. The MRE of ANNs trained with the modified GA-LM algorithm is less than 8.1%, which is smaller than those trained with the existing GA-LM algorithm and the LM algorithm (greater than 15%). Lastly, the ANN synthesis models are validated by the CMT analysis, electromagnetic simulation, and measurements.

• Structural Engineering and Mechanics
• /
• 제50권1호
• /
• pp.37-52
• /
• 2014

The accurate determination of cable tension is important to the monitoring of the condition of a cable-stayed bridge. When applying a vibration-based formula to identify the tension of a real cable under sag, stiffness and boundary conditions, the resulting error must not be overlooked. In this work, by resolving the implicit frequency function of a real cable under the above conditions numerically, indirect methods of determining the cable force and a method to calculate the corresponding cable mode frequency are investigated. The error in the tension is studied by numerical simulation, and an empirical error correction formula is presented by fitting the relationship between the cable force error and cable parameters ${\lambda}^2$ and ${\xi}$. A case study on two real cables of the Shanghai Changjiang Bridge shows that employing the method proposed in this paper can increase the accuracy of the determined cable force and reduce the computing time relative to the time required for the finite element model.

• 한국통신학회논문지
• /
• 제17권9호
• /
• pp.909-917
• /
• 1992

본 논문은 출력 오차 분산(分散)를 갖는 상태공간(狀態空間) 디지털 필터의 실현(實現)을 제안하였다. 이 앨고리즘은 가중함수(加重函數)에 의해서 선형 시불변 시스템내의 가제어성 및 가관측성 gramian을 변화시키는 것이며, 상태공간(常態空間) 계수 변동에 대한 출력 오차 분산(分散)을 줄임으로써 디지털 필터의 성능을 개선할 수 있다. 수치예에서, 본 앨고리즘 구조는 다른 4가지 구조(표준형, 병열형, 통계적감도형, 평형형) 보다 더 적은 출력오차 분산(分散)을 갖는 것을 알았다.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2001년도 가을 학술발표회 논문집
• /
• pp.34-42
• /
• 2001

In this paper, we study on a modified mesh generation method based on the pollution error estimate. This method is designed for the control of the pollution error in any patch of elements of interest. It is a well-known fact that the pollution error estimates are much more than the local one. Reliable a posteriori error estimation is possible by controlling the pollution error in the patch through proper design of the mesh outside the patch. This design is possible by equally distributing the pollution error indicators over the mesh outside the patch. The conventional feedback pollution-adaptive mesh generation algorithm needs many iterations. Therefore, the solution time is significant. But we use the remeshing scheme in the proposed method. We will also show that the pollution error reduces less than the local error.

• Journal of Positioning, Navigation, and Timing
• /
• 제1권1호
• /
• pp.51-57
• /
• 2012

In this paper, design of an error model is presented in which the bias characteristic of the MEMS IMU is taken into consideration for performance enhancement of the MEMS IMU-based GPS/INS integrated navigation system. The drift bias of the MEMS IMU is modeled as a 1st-order Gauss-Markov (GM) process, and the autocorrelation function is obtained from the collected IMU data, and the correlation time is estimated from this. Prior to obtaining the autocorrelation function, the noise of IMU data is eliminated based on wavelet. As a result of simulation, it is represented that the parameters of error model can be estimated correctly only when a proper denoising is performed according to dynamic behavior of drift bias, and that the integrated navigation system based on error model, in which the drift bias is considered, provides more correct navigation performance compared to the integrated navigation system based on error model in which the drift bias is not considered.

• 한국조사연구학회:학술대회논문집
• /
• 한국조사연구학회 2003년도 춘계학술발표대회
• /
• pp.34-41
• /
• 2003

많은 경우, 측정오차분산은 알려지지 않은 참값 또는 참값과 연관된 공변수들의 함수로 표현될 수 있다 이 논문은 단위 당 반복측정에 기초한 단위 내 표본분산을 이용한 선형측정오차분산의 추정에 관한 연구이다 이 논문은 다음의 내용을 포함한다: (1) 측정오차의 크기를 나타내는 상수 $\delta$의 추정; (2) 유한모집단으로부터의 복합표본, 작은 측정오차라는 조건하에 선형측정오차분산의 추정; (3) 부표본에 포함될 확률을 설명하기 위한 성향틴헝 추정 미국의 제3차 건강영양조사자료를 사용하여 이상의 결과들을 이용한 경험적 분석을 실행하였다.

• 한국정밀공학회지
• /
• 제18권4호
• /
• pp.97-103
• /
• 2001

Accuracy of a machined part is determined by the relative motion between the cutting tool and the workpiece. One of the important factors which affects the relative motion is the geometric errors of a machine tool. In this study, firstly, geometric errors are measured by laser interferometer, and the positioning error of each control point selected uniformly on the control surface CAD model can be estimated from th oirm shaping model and geometric error data base. Where a form shaping function is derived from the link of homogeneous transformation matrix. Secondly, control points are shifted to the estimated amount of positioning errors. A new control surface is modeled with NURBS(Non Uniform Rational B-Spline) surface approximation to the shifted control points. By generating tool paths to the redesigned control surface, we reduce the machining error quite.

• 연구논문집
• /
• 통권34호
• /
• pp.101-112
• /
• 2004

In order to analyze the motion errors of the aerostatic stage, it is necessary to consider the influence of the moment variation occurredinside the pads. In this paper, a motion error anaysis method utilizing the transfer functions on the reaction force and moment is proposed, and general characteristics of the transfer functions are discussed. Calculated motion errors by the proposed method show good agreement with the ones calculated by Multi Pad Method, which is considered the entire table as an analysis object. Also, by the introduction of the transfer function of motion errors, which represent the relationship between the spatial frequency components of the rail form error and motion errors, motional characteristics of the porous aerostatic stage can be generalized. In detail, the influence of the spatial frequencies is analyzed quantatively, and the patterns of the insensitive frequencies which almost do not affect the linear motion error or angular motion error according to the rail length ratio and the number of the pad are verified. The relationship between the moment variation occurred inside the pads and the motion errors is also verified together.