• 한국수소및신에너지학회논문집
• /
• 제21권4호
• /
• pp.258-263
• /
• 2010

A numerical stack model has been developed to predict the temperature at a constant-load operation of molten carbonate fuel cell stacks. For the validity of the model, the simulated results with several boundary conditions were compared in the cell temperature data obtained from 75 kW class MCFC stack operation. It was shown that the simulated results with the existing boundary condition, which the stack outlet temperature was fixed at $650^{\circ}C$, didn't match well with the measured data. On the other hand, the stack model with the outlet temperature modified by the outlet manifold temperature measured from the stack under several electric loads was found to explain the measured cell temperature distribution well. The results show that the model can be used to predict the cell temperature distribution in the stacks by the measurement of the manifold outlet temperature.

• 충청수학회지
• /
• 제33권2호
• /
• pp.237-253
• /
• 2020

This article studies asymptotic stability of non-auto nomous linear systems with time-dependent coefficient matrices {A(t)}_t∈ℝ. The classical theorem of Levinson has been widely used to science and engineering non-autonomous systems, but systems with defective eigenvalues could not be covered because such a family does not allow continuous diagonalization. We study systems where the family allows to have upper triangulation and to have defective eigenvalues. In addition to the wider applicability, working with upper triangular matrices in place of Jordan form matrices offers more flexibility. We interpret our and earlier works including Levinson's theorem from the perspective of invariant manifold theory.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송공학회 2015년도 추계학술대회
• /
• pp.4-7
• /
• 2015

In this paper, we propose a matrix completion algorithm for Internet of Things (IoT) localization. The proposed algorithm recovers the Gram matrix of sensors by performing optimization over the Riemannian manifold of fixed-rank positive semidefinite matrices. We compute and show the closed forms of all the differentially geometric components required for applying nonlinear conjugate gradients combined with Armijo line search method. The numerical experiments show that the performance of the proposed algorithm in solving IoT localization is outstanding compared with the state-of-the-art matrix completion algorithms both in noise and noiseless scenarios.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2018년도 추계학술발표대회
• /
• pp.763-766
• /
• 2018

표정은 사람들 사이에서 감정을 표현하는 강력한 비언어적 수단이다. 표정 인식은 기계학습에서 아주 중요한 분야 중에 하나이다. 표정 인식에 사용되는 기계학습 모델들은 사람 수준의 성능을 보여준다. 하지만 좋은 성능에도 불구하고, 기계학습 모델들은 표정 인식 결과에 대한 근거나 설명을 제공해주지 못한다. 이 연구는 표정 인식의 근거로서 Facial Action Coding Unit(AUs)을 사용하기 위해서 CK+ Dataset을 사용하여 표정 인식을 학습한 Convolutional Neural Network(CNN) 모델이 추출한 특징들을 t-distributed stochastic neighbor embedding(t-SNE)을 사용하여 시각화한 뒤, 인식된 표정과 AUs 사이의 분포의 연관성을 확인하는 연구이다.

• East Asian mathematical journal
• /
• 제37권5호
• /
• pp.577-584
• /
• 2021

Let p : X → D be a proper surjective holomorphic submersion where X is a Kähler manifold and D is the unit disc in ℂ. Let Ω be a d-closed semi-positive real (1, 1)-form on X. If each X_s := p^-1(s) for s ∈ D satisfies $-c_1(X_s)+{\Omega}{\mid}_{X_s}$ is Kähler, then the Kähler-Ricci flow twisted by ${\Omega}{\mid}_{X_s}$ has a long time solution by Cao's theorem. This family of twisted Kähler-Ricci flows induces a relative Kähler form ω(t) on the total space X. In this paper, we prove that the positivity of ω(t) is preserved along the twisted Kähler-Ricci flow.

• 대한수학회지
• /
• 제59권2호
• /
• pp.337-352
• /
• 2022

We consider the set of points with historic behavior (which is also called the irregular set) for continuous flows and suspension flows. In this paper under the hypothesis that (X_t)_t is a continuous flow on a d-dimensional Riemaniann closed manifold M (d ≥ 2) with gluing orbit property, we prove that the set of points with historic behavior in a compact and invariant subset ∆ of M is either empty or is a Baire residual subset on ∆. We also prove that the set of points with historic behavior of a suspension flows over a homeomorphism satisfyng the gluing orbit property is either empty or Baire residual and carries full topological entropy.

• Journal of applied mathematics & informatics
• /
• 제34권3_4호
• /
• pp.257-267
• /
• 2016

We find examples of Ideals in a tridiagonal algebra ALGL_∞ and study some properties of Ideals in ALGL_∞. We prove the following theorems: Let k and j be fixed natural numbers. Let A be a subalgebra of ALGL_∞ and let A_2,{k} ⊂ A ⊂ {T ∈ ALGL_∞ | T_(2k-1,2k) = 0}. Then A is an ideal of ALGL_∞ if and only if A = A_2,{k} where A_2,{k} = {T ∈ ALGL_∞ | T_(2k-1,2k) = 0, T_(2k-1,2k-1) = T_(2k,2k) = 0}. Let B be a subalgebra of ALGL_∞ such that B_2,{j} ⊂ B ⊂ {T ∈ ALGL_∞ | T_(2j+1,2j) = 0}. Then B is an ideal of ALGL_∞ if and only if B = B_2,{j}, where B_2,{j} = {T ∈ ALGL_∞ | T_(2j+1,2j) = 0, T_(2j,2j) = T_(2j+1,2j+1) = 0}.

• 태양에너지
• /
• 제9권3호
• /
• pp.37-43
• /
• 1989

This paper dealt with thermal storage efficiency due to difference ($T_{\infty}-Ti$) between the mean temperature of water in the storage tank [$0.5m{\times}0.5m{\times}1.0m$] and the temperature of water flowing into the tank, flow rate of water flowing into the tank and shape of porous manifold which water flow into the tank through. As results of experiments; (1) When the flow rate was constant and the diameter of porous section decreased by 8mm, 6mm, and 4mm, the thermal storage efficiency increased. (2) When the diameter of porous section was constant and the difference ($T_{\infty}-Ti$) between the mean temperature of water in the storage tank and the temperature of water flowing into the tank increased by -30, -20, -10, 5, 10, 15 ($^{\circ}C$), the thermal storage efficiency increased. (3) When the($T_{\infty}-Ti$) was constant and the flow rate decreased by 0.8, 0.4, 0.25(LPM), the thermal storage efficiency increased. (4) When the shape of porous section was rigid, the thermal storage efficiency was the most effective, and with establishing flexible porous section or mesh, the effective thermal storage efficiency was obtained.

• Progress in Superconductivity
• /
• 제12권1호
• /
• pp.46-50
• /
• 2010

In the superconducting $Sr_2VO_3FeAs$, containing bimetallic layers, with maximum $T_c{\approx}\;46\;K$ correlation effects on V ions have been investigated using LDA+U method. Within the local density approximation (LDA) this system has the one-third filled $t_{2g}$ manifold of V, decomposed into $d_{xy}$ of bandwidth W=2 eV and nearly degenerate $d_{zx}d_{yz}$ of W=1 eV. Consideration of correlation effects leads to a metal-insulator transition on V ions $t^{2\uparrow}_{2g}\;{\rightarrow}\;d^{1\uparrow}_{xz}\;d^{1\uparrow}_{yz}$ at the critical on-site Coulomb repulsion $U_c$= 3.5 eV. At U=4 eV, the electronic structure, in which V ions are insulating, leads to several van Hove singularities near $E_F$ and similar Fermiology with other pnictides. Applying U to V ions results in increasing Fe moment as well as V moment, indicating somewhat hybridization between Fe and V ions even though this system is strongly 2-dimesional. Our results show possible importance of correlation effects on this system.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권9호
• /
• pp.2991-3007
• /
• 2022

Two dimensional locality preserving projections (2D-LPP) is an improved algorithm of 2D image to solve the small sample size (SSS) problems which locality preserving projections (LPP) meets. It's able to find the low dimension manifold mapping that not only preserves local information but also detects manifold embedded in original data spaces. However, 2D-LPP is simple and elegant. So, inspired by the comparison experiments between two dimensional linear discriminant analysis (2D-LDA) and linear discriminant analysis (LDA) which indicated that matrix based methods don't always perform better even when training samples are limited, we surmise 2D-LPP may meet the same limitation as 2D-LDA and propose a novel matrix exponential method to enhance the performance of 2D-LPP. 2D-MELPP is equivalent to employing distance diffusion mapping to transform original images into a new space, and margins between labels are broadened, which is beneficial for solving classification problems. Nonetheless, the computational time complexity of 2D-MELPP is extremely high. In this paper, we replace some of matrix multiplications with multiple multiplications to save the memory cost and provide an efficient way for solving 2D-MELPP. We test it on public databases: random 3D data set, ORL, AR face database and Polyu Palmprint database and compare it with other 2D methods like 2D-LDA, 2D-LPP and 1D methods like LPP and exponential locality preserving projections (ELPP), finding it outperforms than others in recognition accuracy. We also compare different dimensions of projection vector and record the cost time on the ORL, AR face database and Polyu Palmprint database. The experiment results above proves that our advanced algorithm has a better performance on 3 independent public databases.