• 대한수학회보
• /
• 제51권5호
• /
• pp.1357-1368
• /
• 2014

This paper is concerned with a reaction-diffusion single species model with harvesting on n-dimensional isotropically growing domain. The model on growing domain is derived and the corresponding comparison principle is proved. The asymptotic behavior of the solution to the problem is obtained by using the method of upper and lower solutions. The results show that the growth of domain takes a positive effect on the asymptotic stability of positive steady state solution while it takes a negative effect on the asymptotic stability of the trivial solution, but the effect of the harvesting rate is opposite. The analytical findings are validated with the numerical simulations.

• 한국산업정보학회논문지
• /
• 제12권4호
• /
• pp.126-130
• /
• 2007

연속 반복되는 시스템에서 제어하는 제어기 repetitive controller를 설계하는 방법과 시스템 안정성을 소개하고자 한다. 제어기는 반복함수로서 수렴을 조장하는 비용함수를 주파수영역에서 최소화함으로써 유도할 수 있다. 모의실험은 제안된 단수모형설계 기법을 이용한 RC를 어떻게 유도하는 것을 보여 주고 있다.

• 대한전자공학회논문지SD
• /
• 제42권4호
• /
• pp.45-52
• /
• 2005

본 논문은 인쇄 회로 기판에 있는 through hole vias를 시간 영역과 주파수 영역 측정을 통하여 characterization을 하였다. Via characterization은 Time Domain Reflectometry (TDR)를 이용하여 시간 영역에서 측정하고 HSPICE fitting 시뮬레이션으로 via 모델 파라미터를 추출하였다. 또한 2 port Vector Network Analyzer (VNA)로 주파수 영역에서 측정하고 Advanced Design System (ADS) fitting 시뮬레이션 하였다. VNA를 이용한 측정에서는 같은 평면에서 probing하기 위해 ABCD matrix 를 이용하여 do-embedding 수식을 유도하였다. 그리고 single via characterization 결과를 바탕으로 differential signaling을 위한 differential via characterization을 TDR과 VNA 측정을 통하여 수행하였다. Differential via characterization은 TDR 모듈의 odd mode와 even mode 소스들을 이용하여 시간 영역에서 측정하고 HSPICE로 fitting 시뮬레이션으로 모델 파라미터를 추출하였다. 추출된 모든 data는 측정 및 simulation 결과를 비교한 결과 single via의 경우, 최대 $14\%$, differential via의 경우 최대 $17\%$의 오차를 나타내었다.

• 한국세라믹학회지
• /
• 제26권1호
• /
• pp.37-42
• /
• 1989

Periodic layered domain structures in doped LiNbO3 crystals grown by Czochralski method were obtained by thermal fluctuation and crystal rotation with inhomogeneous radial temprature distribution. The stressinduced domain formation mechanism model was suggested and discussed.

• Journal of Information Processing Systems
• /
• 제17권5호
• /
• pp.867-878
• /
• 2021

In context awareness and user intention tasks, dataset construction is expensive because specific domain data are required. Although pretraining with a large corpus can effectively resolve the issue of lack of data, it ignores domain knowledge. Herein, we concentrate on data domain knowledge while addressing data scarcity and accordingly propose a multi-channel long short-term memory (LSTM). Because multi-channel LSTM integrates pretrained vectors such as task and general knowledge, it effectively prevents catastrophic forgetting between vectors of task and general knowledge to represent the context as a set of features. To evaluate the proposed model with reference to the baseline model, which is a single-channel LSTM, we performed two tasks: voice phishing with context awareness and movie review sentiment classification. The results verified that multi-channel LSTM outperforms single-channel LSTM in both tasks. We further experimented on different multi-channel LSTMs depending on the domain and data size of general knowledge in the model and confirmed that the effect of multi-channel LSTM integrating the two types of knowledge from downstream task data and raw data to overcome the lack of data.

• 한국멀티미디어학회논문지
• /
• 제25권10호
• /
• pp.1375-1385
• /
• 2022

In this paper, we propose a semi-supervised domain adaptation solution to deal with practical face recognition (FR) scenarios where a single face image for each target identity (to be recognized) is only available in the training phase. Main goal of the proposed method is to reduce the discrepancy between the target and the source domain face images, which ultimately improves FR performances. The proposed method is based on the Domain Adatation network (DAN) using an MMD loss function to reduce the discrepancy between domains. In order to train more effectively, we develop a novel loss function learning strategy in which MMD loss and cross-entropy loss functions are adopted by using different weights according to the progress of each epoch during the learning. The proposed weight adoptation focuses on the training of the source domain in the initial learning phase to learn facial feature information such as eyes, nose, and mouth. After the initial learning is completed, the resulting feature information is used to training a deep network using the target domain images. To evaluate the effectiveness of the proposed method, FR performances were evaluated with pretrained model trained only with CASIA-webface (source images) and fine-tuned model trained only with FERET's gallery (target images) under the same FR scenarios. The experimental results showed that the proposed semi-supervised domain adaptation can be improved by 24.78% compared to the pre-trained model and 28.42% compared to the fine-tuned model. In addition, the proposed method outperformed other state-of-the-arts domain adaptation approaches by 9.41%.

• Journal of Power Electronics
• /
• 제12권4호
• /
• pp.623-631
• /
• 2012

In this paper, a digital controller design procedure is presented for single-phase voltage-doubler boost rectifiers (VDBR). The model derivation of the single-phase VDBR is performed in the s-domain. After that the simplified equivalent z-domain models are derived. These z-domain models are utilized to design the input current and the output dc-link voltage controllers. For the controller design in the z-domain, the traditional K-factor method is modified by considering the nature of the digital controller. The frequency pre-warping and anti-windup techniques are adapted for the controller design. By using the proposed method, the phase margin and the control bandwidth are accurately achieved as required by controller designers in a practical frequency range. The proposed method is applied to a 2.5 kVA single-phase VDBR for Uninterruptible Power Supply (UPS) applications. From the simulation and the experimental results, the effectiveness of the proposed design method has been verified.

• Journal of Magnetics
• /
• 제6권1호
• /
• pp.36-41
• /
• 2001

A micromagnetic model and a single-domain model simulation programs were used to analyze the sensitivity of a $20\mu m\times 60\mu m \times 1000{\AA}$ permalloy strip as a magnetoresistance sensor with bias fields of various directions and magnitudes. The micromagnetic model agrees with the measured sensitivity data better than the single-domain model. The data show the highest peak sensitivity with the bias field at 90$^{\circ}$to the current. The peak sensitivity decreases and the peak broadens as the bias angle decreases. The simulation using the micromagnetic model shows that a bias angle smaller than 90$^{\circ}$eads to magnetization patterns which are free from closure domains or vertices over a wider range of bias fields.

• ETRI Journal
• /
• 제38권3호
• /
• pp.487-493
• /
• 2016

Two new methods are proposed for an unsupervised adaptation of a language model (LM) with a single sentence for automatic transcription tasks. At the training phase, training documents are clustered by a method known as Latent Dirichlet allocation (LDA), and then a domain-specific LM is trained for each cluster. At the test phase, an adapted LM is presented as a linear mixture of the now trained domain-specific LMs. Unlike previous adaptation methods, the proposed methods fully utilize a trained LDA model for the estimation of weight values, which are then to be assigned to the now trained domain-specific LMs; therefore, the clustering and weight-estimation algorithms of the trained LDA model are reliable. For the continuous speech recognition benchmark tests, the proposed methods outperform other unsupervised LM adaptation methods based on latent semantic analysis, non-negative matrix factorization, and LDA with n-gram counting.

• 한국전기전자재료학회:학술대회논문집
• /
• 한국전기전자재료학회 2001년도 하계학술대회 논문집
• /
• pp.1050-1053
• /
• 2001

There is great interest in developing magnetoresistance(MR) sensor, using ferromagnetic, electrically non-magnetic conducting and antiferromagnetic films, especially for the use in weak magnetic fields. Here, we report single and Wheatstone-bridge type of MR sensors made in Si(001)/HiO(300$\AA$)/NiFe bilayers. Angular dependence of MR profiles was measured in Si(001)/NiO(300$\AA$)/NiFe(450$\AA$) films as a function of an angle between current and applied field direction, also, linearity was determined. AMR characteristics of single MR sensors was well explained with single domain model. Good linearity in 45$^{\circ}$Wheatstone-bridge type of MR sensors consisting of 4 single MR sensors made in Si(001)/NiO(300$\AA$)/NiFe(450$\AA$) was shown in the range of about $\pm$50 Oe.