• Journal of Biosystems Engineering
• /
• 제12권3호
• /
• pp.17-29
• /
• 1987

A 10-degree of freedom mathematical model of motion for power tiller-trailer system was developed. This model can predict motion characteristics of power tiller trailer system while travelling over smooth and irregular ground surfaces under various operating conditions. The model provide, the fundamental data needed to improve the stability of power tiller-trailer systems.

• International Journal of Automotive Technology
• /
• 제7권7호
• /
• pp.833-839
• /
• 2006

In this paper, roll behavior of three planar half car models are compared. The first model is a simple model whose contact point between a wheel and the ground is assumed to be fixed with a revolute joint. The second model is a modified model of the fIrst model, whose wheel tread width can vary. In this model, the instant center of a wheel with respect to the ground, which is crucial to find the roll center, is assumed to be at the contact point of a wheel and the ground. The last model uses the pole of a wheel with respect to the ground for small displacement as the instant center of a wheel with respect to the ground. Loci of the center of gravity point, the fixed and the moving centrodes which are traces of roll center position in the ground and the body frame respectively, wheel contact points, and instant centers of a wheel with respect to the ground are calculated.

• 전자통신동향분석
• /
• 제33권5호
• /
• pp.103-110
• /
• 2018

Studies on artificial intelligence have been developed for the past couple of decades. After a few periods of prosperity and recession, a new machine learning method, so-called Deep Learning, has been introduced. This is the result of high-quality big- data, an increase in computing power, and the development of new algorithms. The main targets for deep learning are 1D audio and 2D images. The application domain is being extended from a discriminative model, such as classification/segmentation, to a generative model. Currently, deep learning is used for processing 3D data. However, unlike 2D, it is not easy to acquire 3D learning data. Although low-cost 3D data acquisition sensors have become more popular owing to advances in 3D vision technology, the generation/acquisition of 3D data remains a very difficult problem. Moreover, it is not easy to directly apply an existing network model, such as a convolution network, owing to the variety of 3D data representations. In this paper, we summarize the 3D deep learning technology that have started to be developed within the last 2 years.

• 전자통신동향분석
• /
• 제33권4호
• /
• pp.1-13
• /
• 2018

Deep learning (DL) models have been widely applied to AI applications such image recognition and language translation with big data. Recently, DL models have becomes larger and more complicated, and have merged together. For the accelerated training of a large-scale deep learning model, model parallelism that partitions the model parameters for non-shared parallel access and updates across multiple machines was provided by a few distributed deep learning frameworks. Model parallelism as a training acceleration method, however, is not as commonly used as data parallelism owing to the difficulty of efficient model parallelism. This paper provides a comprehensive survey of the state of the art in model parallelism by comparing the implementation technologies in several deep learning frameworks that support model parallelism, and suggests a future research directions for improving model parallelism technology.

• 한국전산구조공학회논문집
• /
• 제35권1호
• /
• pp.1-8
• /
• 2022

본 연구에서는 토양-탄체 간 분리 및 재접촉을 고려한 IFL 기반 침투해석 기술을 개발하고 이를 기존 문헌의 실험결과와 비교하는 연구를 수행했다. 탄체를 강체로 가정한 후, 토양 내로 침투 시 발생하는 구형공동팽창 현상을 고려함으로써 탄체의 궤적을 예측할 수 있다. 토양에 대한 저항함수는 Mohr-Coulomb 항복 모델을 활용했으며, 입사각 혹은 AOA에 따른 J-hook 현상을 모사할 수 있다. 기존 문헌에서의 실험결과(총 6회)와의 비교 결과, 수치해석으로부터 예측한 탄체의 침투 깊이는 실험대비 약 13.4%의 평균오차를 나타냈다. 일반적으로 탄체의 침투 경로를 예측하기 위해 유한요소법이 널리 활용된다. 하지만, 유한요소법 활용 시, 탄체의 모델링을 위해 많은 시간과 노력이 필요하며, 해석 수행을 위해 수 시간이 소요된다. 본 연구를 통해 개발한 모델을 활용할 시, 탄체의 치수 입력만 필요하며 해석 시간도 수 초 이내이다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2002년도 추계학술대회 논문집 전기기기 및 에너지변환시스템부문
• /
• pp.286-293
• /
• 2002

Photovoltaic systems normally use a maximum power point tracking (MPPT) technique to continuously deliver the highest possible power to the load when variations in the insolation and temperature occur. A simple method of tracking the maximum power points (MPPs) and forcing the boost converter system to operate close to these Points is presented through deriving small-signal model and transfer function of boost converter. This paper aims at modeling boost converter including equivalent series resistance of input reservoir capacitor by state-space-averaging method and PWM switch model. In the future, properly designed controller for compensation will be constructed in real system for maximum photovoltaic power tracking control.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2003년도 하계학술대회 논문집 A
• /
• pp.83-85
• /
• 2003

This paper presents a correction technique of missing load data. In this paper, the ARIMA(Autoregressive Integrated Moving Average) model and Piecewise Cubic Interpolation are applied to seek the missing parameters. The new model has been tested under a variety of conditions and it is shown in this paper to produce excellent results. It is helpful for operators to designed the load duration curve.

• 한국공작기계학회논문집
• /
• 제16권4호
• /
• pp.87-92
• /
• 2007

The full automation welding has not yet been achieved partly because the mathematical model for the process parameters of a given welding task is not fully understood and quantified. Several mathematical models to control welding quality, productivity, microstructure and weld properties in arc welding processes have been studied. However, it is not an easy task to apply them to the various practical situations because the relationship between the process parameters and the bead geometry is non-linear and also they are usually dependent on the specific experimental results. Practically, it is difficult, but important to know how to establish a mathematical model that can predict the result of the actual welding process and how to select the optimum welding condition under a certain constraint. In this paper, an attempt has been made to develop an neural network model to predict the weld top-bead height as a function of key process parameters in the welding. and to compare the developed models using three different training algorithms in order to select an adequate neural network model for prediction of top-bead height.

• 한국유통학회지:유통연구
• /
• 제15권4호
• /
• pp.61-85
• /
• 2010

본 연구에서는 국내 대형할인점의 확산을 효과적으로 설명하기 위해 기업의 정보와 구매자의 구전으로 확산을 설명하는 Bass모형에 제3의 요소로 공간적 영향력을 고려하였다. 국내 대형할인점의 확산은 확산중심지인 서울경인지역에서 저차중심지인 4개 지역권역으로 확산되는 형태를 보임에 따라 공간적 영향이 중요하게 작용할 것으로 기대된다. 본 연구에서 공간적으로 구분된 시장 A(확산중심지)가 시장 B(저차중심지)에 미치는 영향이 완전히 통제되지 못하는 상황에서 시장 A가 시장 B에 미치는 공간적 영향을 다국가확산모형(multinational diffusion model)을 확장한 공간확산모형(spatial diffusion model)을 이용하여 정의하였다. Bass모형과 공간확산모형의 모수추정을 통해 두 가지 정보전달경로와 관련된 혁신계수와 모방계수로 확산을 설명하는 Bass모형보다 공간확산모형이 국내 대형할인점 확산을 더욱 효과적으로 설명하는 것으로 나타났다. 또한 혁신중심지인 서울경인과 4개 지역권역의 소매환경을 나타내는 개념적 거리에 따라 공간확산모형에서 공간적요인의 영향력이 달라질 것이 기대되어 공간확산계수와 소매환경변수간의 상관관계를 살펴보았고, 연구결과 확산중심지에서 저차중심지에 대한 공간적 영향력은 저차중심지의 소매환경이 확산중심지의 소매환경과 유사할수록 크다는 것을 밝혀내었다.

• Structural Engineering and Mechanics
• /
• 제17권1호
• /
• pp.51-68
• /
• 2004

Since the linear elastic fracture analysis has been proved to be insufficient in predicting the failure of strain hardening materials, a number of fracture concepts have been studied which remain applicable in the presence of plasticity near a crack tip. This work thereby presents a new finite element model to predict the elastic-plastic crack-tip field and fatigue life of center-cracked panels(CCP) with ductile fracture under large-scale yielding conditions. Also, this study has been carried out to investigate the path-dependence of J-integral within the plastic zone for elastic-perfectly plastic, bilinear elastic-plastic, and nonlinear elastic-plastic materials. Based on the incremental theory of plasticity, the p-version finite element is employed to account for the accurate values of J-integral, the most dominant fracture parameter, and the shape of plastic zone near a crack tip by using the J-integral method. To predict the fatigue life, the conventional Paris law has been modified by substituting the range of J-value denoted by ${\Delta}J$ for ${\Delta}K$. The experimental fatigue test is conducted with five CCP specimens to validate the accuracy of the proposed model. It is noted that the relationship between the crack length a and ${\Delta}K$ in LEFM analysis shows a strong linearity, on the other hand, the nonlinear relationship between a and ${\Delta}J$ is detected in EPFM analysis. Therefore, this trend will be depended especially in the case of large scale yielding. The numerical results by the proposed model are compared with the theoretical solutions in literatures, experimental results, and the numerical solutions by the conventional h-version of the finite element method.