• 한국소성가공학회:학술대회논문집
• /
• 한국소성가공학회 2005년도 추계학술대회 논문집
• /
• pp.476-479
• /
• 2005

In this study, the process parameters in powder compaction are optimized for getting high relative densities. To find optimized parameters, the analytic models of powder compaction are firstly prepared by 2-dimensional rod arrays with random green densities using a quasi-random multi-particle array. Then, using finite element method, the changes in relative densities are analyzed by varying the size of the particle, the amplitude of cyclic compaction, and the coefficient of friction, which influence the relative density in cyclic compactions. After the analytic function of relative density associated process parameters are formulated by aid of the response surface method, the optimal conditions in powder compaction process are found by the grid search method.

• 호남수학학술지
• /
• 제27권2호
• /
• pp.301-315
• /
• 2005

Let $\{A_u,\;u=0,\;1,\;2,\;{\cdots}\}$ be a sequence of coefficient matrices such that ${\sum}_{u=0}^{\infty}{\parallel}A_u{\parallel}<{\infty}$ and ${\sum}_{u=0}^{\infty}\;A_u{\neq}O_{m{\times}m}$, where for any $m{\times}m(m{\geq}1)$, matrix $A=(a_{ij})$, ${\parallel}A{\parallel}={\sum}_{i=1}^m{\sum}_{j=1}^m{\mid}a_{ij}{\mid}$ and $O_{m{\times}m}$ denotes the $m{\times}m$ zero matrix. In this paper, a functional central limit theorem is derived for a stationary m-dimensional linear process ${\mathbb{X}}_t$ of the form ${\mathbb{X}_t}={\sum}_{u=0}^{\infty}A_u{\mathbb{Z}_{t-u}}$, where $\{\mathbb{Z}_t,\;t=0,\;{\pm}1,\;{\pm}2,\;{\cdots}\}$ is a stationary sequence of linearly positive quadrant dependent m-dimensional random vectors with $E({\mathbb{Z}_t})={{\mathbb{O}}$ and $E{\parallel}{\mathbb{Z}_t}{\parallel}^2<{\infty}$.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2003년도 ISIS 2003
• /
• pp.599-602
• /
• 2003

A new technique is presented to construct predictive models of plasma etch processes. This was accomplished by combining a backpropagation neural network (BPNN) and a random generator (RC). The RG played a critical role to control neuron gradients in the hidden layer, The predictive model constructed in this way is referred to as a randomized BPNN (RG-BPNN). The proposed scheme was evaluated with a set of experimental plasma etch process data. The etch process was characterized by a 2$^3$ full factorial experiment. The etch responses modeled are 4, including aluminum (Al) etch rate, profile angle, Al selectivity, and do bias. Additional test data were prepared to evaluate model appropriateness. The performance of RC-BPNN was evaluated as a function of the number of hidden neurons and the range of gradient. for given range and hidden neurons, 100 sets of random neuron gradients were generated and among them one best set was selected for evaluation. Compared to the conventional BPNN, the proposed RC-BPNN demonstrated about 50％ improvements in all comparisons. This illustrates that the RG-BPNN of multi-valued gradients is an effective way to considerably improve the predictive ability of current BPNN of single-valued gradient.

• 동력기계공학회지
• /
• 제4권2호
• /
• pp.40-45
• /
• 2000

Some engineering materials are often known to have considerable spatial variation in their resisting strength and other properties. The objective of this study is to investigate the averaging effect and the applicability of extremal statistic for the statistical size effect. In the present study, it is assumed that the material property is a stationary random process in space. The theoretical autocorrelation function of the material strength are discussed for several correlation lengths. And, in order to investigate the statistical size effect, the material properties was simulated by using the non-Gaussian random process method. The material properties were plotted on the Weibull probability papers. The main results are summarized as follows: The autocorrelation function of the material properties are almost independent of the averaging length. The variance decreases with increasing the averaging length. As correlation length is smaller, the slope is larger. And also, it was found that Weibull statistics based on the weakest-link model could not explain the spatial variation of material properties with respect to the size effect satisfactory.

• 한국해양공학회지
• /
• 제17권2호
• /
• pp.54-59
• /
• 2003

The characteristics of the parameters for the probability distribution of fatigue crack growth life, using the non-Gaussian random process simulation method is investigated. In this paper, the material resistance to fatigue crack growth is treated as a spatial random process, which varies randomly on the crack surface. Using the previous experimental data, the crack length equals the number of cycle curves that are simulated. The results are obtained for constant stress intensity factor range conditions with stress ratios of R=0.2, three specimen thickness of 6, 12 and 18mm, and the four stress intensity level. The probability distribution function of fatigue crack growth life seems to follow the 3-parameter Wiubull,, showing a slight dependence on specimen thickness and stress intensity level. The shape parameter, $\alpha$, does not show the dependency of thickness and stress intensity level, but the scale parameter, $\beta$, and location parameter, ${\gamma}$, are decreased by increasing the specimen thickness and stress intensity level. The slope for the stress intensity level is larger than the specimen thickness.

• 대한조선학회논문집
• /
• 제47권6호
• /
• pp.813-820
• /
• 2010

The spread mooring system for FPSO is developed to explore deep sea area, in which swell is dominant. It is known that the tension response of mooring lines in this sea area shows bimodal spectrum. Assuming normal distribution of tension profile and Rayleigh distribution of tension amplitude, the power spectral density function (PSD) of the mooring tension under the bimodal stationary random process is applied for the calculation of spectrum fatigue. Three popular methods, which are simple summation method, combined spectrum method and Jioa-Moan method, are used to combine fatigue damages from bimodal spectrum characteristics. Each damage value is compared with damage using Rainflow Cycle Counting (RCC) method which is believed to be close to exact solution. Vanmarcke' parameter and RMS(Root Mean Square) ratio are employed to assess relative damage variations between from RCC method and from three combination methods. Finally the most reliable fatigue damage combining method for spread mooring system is suggested.

• 한국해양공학회:학술대회논문집
• /
• 한국해양공학회 2002년도 추계학술대회 논문집
• /
• pp.301-306
• /
• 2002

The characteristics of parameters for the probability distribution of fatigue crack growth lives by the non-Gaussian random process simulation method is investigated. In this paper, the material resistance to fatigue crack growth is treated as a spatial random process, which varies randomly on the crack surface. Using the previous experimental data, the crack length - the number of cycles curves are simulated. The results are obtained for constant stress intensity factor range conditions with stress ratio of R=0.2, three specimen thickness of 6, 12 and 18mm, and the four stress intensity level. The probability distribution function of fatigue crack growth lives seems to follow the 3-parameter Wiubull and shows a slight dependence on specimen thickness and stress intensity level. The shape parameter, ${\alpha}$, does not show the dependency of thickness and stress intensity level, but the scale parameter, ${\beta}$, and location parameter, ${\upsilon}$, are decreased by increasing the specimen thickness and stress intensity level. The slope for the stress intensity level is larger than the specimen thickness.

• 전기의세계
• /
• 제17권1호
• /
• pp.48-50
• /
• 1968

The convergence rate of Bayes' learning process is investigated for a binomial random variable. A measure of the rate of convergence is proposed and it is found that such a measure can be approximated by an exponential function of the number of observations.

• 멤브레인
• /
• 제16권3호
• /
• pp.221-229
• /
• 2006

본 연구는 불소관능기인 perfluorocyclobutane (PFCB), fluorenyl, sulfonyl계 방향족 화합물을 동시에 포함하는 술폰화된 랜덤 고분자 전해질 막의 제조 및 그 특성에 관한 것이다. 이러한 고분자 전해질 막은 세단계의 합성, 즉 trifluorovinyloxy그룹을 양말단에 포함하는 단량체의 합성, 중부가반응 형태의 열중합, 그리고 chlorosulfonic acid를 이용한 후술폰화를 통하여 얻어졌다. 후술폰화 반응은 고분자 내에 포함된 fluorenyl기의 함량에 따라 일정한 몰비의 술폰화제 비율로 고정하여 진행되었으며, 이에 따라 다양한 이온교환 능력(IEC)을 가지는 고분자를 쉽게 합성할 수 있었다. 제조된 단량체 및 고분자들의 구조와 순도는 각각 FT-IR과 NMR 그리고 질량분석기를 통하여 확인되었다. 술폰화된 fluorenyl기가 많아질수록 술폰화도와 이온교환 능력이 증가하는 것을 확인할 수 있었고 그에 따른 함수율도 역시 증가하는 거동을 보였다. 술폰화된 고분자들의 이온전도도를 측정한 결과 술폰화도가 증가할수록 이온 전도도가 증가하는 것을 관찰할 수 있었다, 이렇게 제조된 전해질막 중 S-1과 S-2의 경우 다양한 온도범위($25{\sim}80^{\circ}C$)에서 상용 전해질막인 Nafion-115를 능가하는 우수한 이온전도도를 나타냈다.

• 한국산림과학회지
• /
• 제106권2호
• /
• pp.213-218
• /
• 2017

본 연구의 목적은 임의효과(random effect)를 이용하여 충남지역 임령-바이오매스 모형을 개발하고 임의효과의 적용성을 평가하는데 있다. 충남지역 소나무림의 임령에 따른 바이오매스 모형 개발을 위해 임분 구조를 고려하여 전국의 중부지방소나무 임분에서 30개소(150그루)를 조사하고 임령과 바이오매스 자료를 수집하였다. 모형 개발에서 중부지방소나무의 임령-바이오매스 관계는 고정효과(fixed effect)이고 지역간 차이를 임의효과로 설정하였다. 임의효과에 따른 모형의 적합도를 검정하기 위해 아카이케의 정보기준(Akaike Information Criterion, AIC)을 참고하고 지역간 차이에 따른 분산-공분산 행렬과 오차항을 추정하였다. 추정된 공분산은 -1.0022, 오차항은 0.6240이고 분산-공분산 행렬을 이용한 임의효과 모형의 AIC는 377.7을 나타내어 선행 연구와 이질적인 차이는 없었다. 이러한 결과는 범주형 자료의 임의효과가 모형 개발에 반영된 결과로 판단된다. 본 연구의 결과는 임의효과를 이용하여 일부지역에 국한되어 개발되었던 바이오매스 모형 연구에 활용이 가능하다.