• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2010.06a
• /
• pp.307-307
• /
• 2010

최근 유한연료의 고갈로 인해 세계 유가가 불안정 됨으로서 대체 에너지에 대한 연구가 많이 진행 되고 있다. 특히 압전 소자를 이용한 에너지 하베스팅은 압전 역효과를 이용한 것으로서 주변에서 무의미하게 버려지는 진동이나 바람, 열 에너지를 실 생활에 사용할 수 있는 전기 에너지로 변환할 수 있는 유망한 기술 중 하나이다. 이러한 에너지 하베스팅 기술은 일본과 같은 선진국에서 이미 지하철 및 일반 다리와 같이 진동이 극히 많은 곳에서 응용되고 있다. 이러한 에너지 하베스팅 기술을 응용 하려면 전압출력 계수($g_{33}$)가 높아야 한다. 이것은 압전 d 상수와 유전상수에 영향을 많이 받는 것으로 알려져 있다. 현재가지 응용되는 압전 하베스팅 조성은 Pb(Zr,Ti)$O_3$ (PZT)를 기초로한 세라믹이 응용되고 있다. Pb(Zr,Ti)$O_3$ (PZT) 세라믹은 Morpohotropic phase boundary(MPB)에서 전기기계 결합계수 (kp) 와 기계적 품질계수 (Qm) 이 각각 0.5와 500으로 우수한 특성을 나타낸다. 또한 큐리온도 (Tc) 도 $400^{\circ}C$로 온도 안정성 또한 높다. 하지만 $1000^{\circ}C$ 이상에서 소결하는 PbO는 소결 중 급격한 휘발로 환경적 오염 뿐 아니라 특성의 저하를 야기시킨다. 그래서 몇몇 나라에서는 그 사용을 제한하고 점차적으로 사용을 줄여 나가고 있는 동시에 PbO가 첨가되어 있지 않은 Lead-Free 세라믹의 연구가 많이 진행되고 있다. Lead-Free 세라믹 중 alkaline niobate를 기초로 한 페로브스카이트 구조의 ($Na_{0.5}K_{0.5})NbO_3$ (NKN) 은 PbO를 기초로 한 세라믹을 대체할 유망한 후보자 중 하나이다. 하지만 NKN세라믹의 K 성분의 조해성 및 고온에서의 휘발로 인해 일반 적인 소결 방법으로는 고밀도의 세라믹을 얻기 매우 어렵다. 그래서 Hot pressing, Hot forging, RTGG(Reactive Template Grain Growth), SPS(Spark plasma Sintering)와 같은 특별한 소결 법을 이용하거나 $K_8CuNb_4O_{23}$(KCN) 이나 $K_{5.4}Cu_{1.3}Ta_{10}O_{29}$(KCT) 등을 첨가하여 그 소결성을 향상 시키는 방법도 있다. 또한 압전 d상수를 향상 시키기 위해 $Nb_2O_5$나, $La_2CO_3$, $CeO_2$, $Li_2CO_3$ 등을 치환함으로써 압전 d상수를 향상 시켜 전압출력 계수를 높이는 연구 또한 많은 보고가 되어 있다. 특히 $Li_2CO_3$의 첨가는 일반 적인 소결 방법으로도 밀도의 조밀함을 향상 시켜 그에 따른 높은 유전율과 전기기계 결합계수, 압전 d상수를 가져 많은 연구가 되어지고 있다. 그래서 본 연구에서는 일반적인 ($K_{0.5}N_{0.5})_{1-x}Li_x(Nb_{0.96}Sb_{0.04})O_3$ + 0.2mol%$La_2O_3$ + 1.2mol%$K_8CuNb_4O_{23}$ 세라믹에 x(=Li) 치환에 따른 유전 및 압전특성을 조사하였다.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.48 no.1
• /
• pp.77-83
• /
• 2011

In this thesis, 50W Doherty amplifier was designed and implemented for Beyond 3G's repeater and base-station. Auxiliary amplifier of doherty amplifier was implemented by Gate bias control circuit. Though gate bias control circuit solved auxiliary's bias problem, output characteristics of doherty amplifier was limited. To enhance the output characteristic relativize Drain control circuit And To improve power efficiency make 3-way Doherty power amplifier. therefore, 3-way GDCD (Gate and Drain bias Control Doherty) power amplifier is embodied to drain bias circuit for General Doherty power amplifier. The 3-way GDCD power amplifier composed of matching circuit with chip capacitor and micro strip line using FR4 dielectric substance of specific inductive capacity(${\varepsilon}r$) 4.6, dielectric substance height(H) 30 Mills, and 2.68 Mills(2 oz) of copper plate thickness(T). Experiment result satisfied specification of amplifier with gains are 57.03 dB in 2.11 ~ 2.17 GHz, 3GPP frequency band, PEP output is 50.30 dBm, W-CDMA average power is 47.01 dBm, and ACLR characteristics at 5MHz offset frequency band station is -40.45 dBc. Especially, 3-way DCHD power amplifier showed excellence efficiency performance improvement in same ACLR than general doherty power amplifier.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.49 no.12
• /
• pp.209-218
• /
• 2012

Recently, the Field Programmable Stateful Logic Array (FPSLA) was proposed as one of the most promising system integration technologies which will extend the life of the Moore's law. This work is the first proposal of the FPSLA design automation flow, and the approaches to logic synthesis, synchronization, physical mapping, and automatic placement of the FPSLA designs. The synchronization at each gate for pipelining determines the x-coordinates of cells, and reduces the placement to 1-dimensional problems. The objective function and its gradients for the non-linear optimization of the net length and placement density have been remodeled for the reduced global placement problem. Also, a recursive algorithm has been proposed to legalize the placement by relaxing the density overflow of bipartite bin groups in a top-down hierarchical fashion. The proposed model and algorithm are implemented, and validated by applying them to the ACM/SIGDA benchmark designs. The output state of a gate in an FPSLA needs to be duplicated so that each fanout gate can be connected to a dedicated copy. This property has been taken into account by merging the duplicated nets into a hyperedge, and then, splitting the hyperedge into edges as the optimization progresses. This yields additional 18.4% of the cell count reduction in the most dense logic stage. The practicality of the FPSLA can be further enhanced primarily by incorporating into the logic synthesis the constraint to avoid the concentrated fains of gates on some logic stages. In addition, an efficient algorithm needs to be devised for the routing problem which is based on a complicated graph. The graph models the nanowire crossbar which is trimmed to be embedded into the FPSLA fabric, and therefore, asymmetric. These CAD tools can be used to evaluate the fabric efficiency during the architecture enhancement as well as automate the design.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 1995.10a
• /
• pp.289-295
• /
• 1995

구조물의 동적부하에 대한 동적변형 응답을 정확히 예측하고, Over Design이나 Under Design이 아닌 합리적인 설계방안의 개발은 중요한 과제이다. 동적강도해석이나 소음 승차감과 같은 진동 및 충격에 기인하는 제반 문제를 복잡한 구조물을 대상으로 합리적으로 처리하기 위한 Dynamic Design Analysis는 높은 신뢰성의 추구와 더불어 필요불가결한 기술이 되고 있다. 동적해석 방법으로는 현재 유한요소법이 널리 사용되고 있으며 여러 종류의 범용 프로그램들이 보급되어 있는 실정이다. 그러나 특히 동적문제에 있어서는 형상이나 거동이 복잡한 구조물의 경우, 또는 차량의 차체와 같이 많은 장착물이 부착된 경우에는 유한요소법의 적용이 곤란하여, 지금까지 대처할 수 있는 유용한 방법이 없었다. 따라서 비교적 용이하고 간단하게 적용가능한 진동실험을 기초로 한 구조물의 동적 응답해석 및 설계 방안의 개발이 필요하다. 본 연구에서는 진동시험으로 얻어진 부분구조물의 응답특성과 결합특성으로부터 결합 후의 응답특성을 예측할 수 있는 방법을 전달함수합성이론을 기초로하여 프로그래밍 package화 한다. 그리고 평판구조물에 대하여 진동시험과 컴퓨터 시뮬레이션을 통하여 개발된 방법의 타당성을 검증한다. 또한 실제 차량에서 차체만의 진동시험과 엔진의 자유진동시험에 의한 시험데이터로부터 차체와 엔진이 마운트 결합된 후의 진동특성을 예측한다. 진동시험시에 입력과 출력에 노이즈가 필연적으로 혼입되어 주파수응답함수의 크기(magnitude)와 위상(phase)을 왜곡시킨다. 특히 위상의 왜곡은 복소수연산을 하는 전달함수합성법의 결과에 중요한 영향을 미치게 된다. 본 연구에서는 데이타 획득시 입력과 출력의 시간지연으로 생기는 위상왜곡을 보정하는 방법을 제시하고, 그 개선 정도를 조사한다.는 소견의 확실도로서 가능성을 표현한 것이다. 예를 들면, 진동진폭 스펙트럼상에 2X 성분이 상당히 크게 나타나 정렬불량의 가능성이 0.7 정도라고 판정하는 것 등은 이러한 수치적진리치를 이용하는 방법이다. 그러나 상기의 수치적 표현만으로는 확실도를 한개의 수치로서 대표하게 하는 것은 진단의 정밀도에 문제가 있을 것으로 생각된다. 따라서 언어적진리치가 도입되어 [상당히 확실], [확실], [약간 확실] 등의 언어적인 표현을 이용하여 애매성을 표현하게 되었다. 본 논문에서는 간이진단 결과로부터 추출된 애매한 진단결과중에서 가장 가능성이 높은 이상원인을 복수로 선정하고, 여러 종류의 수치화할 수 없는 언어적(linguistic)인 정보ㄷㄹ을 if-then 형식의 퍼지추론으로 종합하는 회전기계의 이상진단을 위한 정밀진단 알고리즘을 제안하고 그 유용성을 검토한다. 존재하여도 모우드 변수들을 항상 정확하게 구할 수 있으며, 또한 알고리즘의 안정성이 보장된 것이다.. 여기서는 실험실 수준의 평 판모델을 제작하고 실제 현장에서 이루어질 수 있는 진동제어 구조물에 대 한 동적실험 및 FRS를 수행하는 과정과 동일하게 따름으로써 실제 발생할 수 있는 오차나 error를 실험실내의 차원에서 파악하여 진동원을 있는 구조 물에 대한 진동제어기술을 보유하고자 한다. 이용한 해마의 부피측정은 해마경화증 환자의 진단에 있어 육안적인 MR 진단이 어려운 제한된 경우에만 실제적 도움을 줄 수 있는 보조적인 방법으로 생각된다.ofile whereas relaxivity at high field is not affected by τS. On the other hand, the change in τV does not affect low field profile but strongly in fluences on bot

• The Korean Journal of Applied Statistics
• /
• v.32 no.5
• /
• pp.693-702
• /
• 2019

Deep learning has gained popularity for the classification and prediction task. Neural network layers become deeper as more data becomes available. Saturation is the phenomenon that the gradient of an activation function gets closer to 0 and can happen when the value of weight is too big. Increased importance has been placed on the issue of saturation which limits the ability of weight to learn. To resolve this problem, Glorot and Bengio (Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, 249-256, 2010) claimed that efficient neural network training is possible when data flows variously between layers. They argued that variance over the output of each layer and variance over input of each layer are equal. They proposed a method of initialization that the variance of the output of each layer and the variance of the input should be the same. In this paper, we propose a new method of establishing initialization by adopting truncated normal distribution and truncated cauchy distribution. We decide where to truncate the distribution while adapting the initialization method by Glorot and Bengio (2010). Variances are made over output and input equal that are then accomplished by setting variances equal to the variance of truncated distribution. It manipulates the distribution so that the initial values of weights would not grow so large and with values that simultaneously get close to zero. To compare the performance of our proposed method with existing methods, we conducted experiments on MNIST and CIFAR-10 data using DNN and CNN. Our proposed method outperformed existing methods in terms of accuracy.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.42 no.5 s.305
• /
• pp.123-136
• /
• 2005

We propose a recognition system for superimposed patterns based on selective attention model and SVM which produces better performance than artificial neural network. The proposed selective attention model includes attention layer prior to SVM which affects SVM's input parameters. It also behaves as selective filter. The philosophy behind selective attention model is to find the stopping criteria to stop training and also defines the confidence measure of the selective attention's outcome. Support vector represents the other surrounding sample vectors. The support vector closest to the initial input vector in consideration is chosen. Minimal euclidean distance between the modified input vector based on selective attention and the chosen support vector defines the stopping criteria. It is difficult to define the confidence measure of selective attention if we apply common selective attention model, A new way of doffing the confidence measure can be set under the constraint that each modified input pixel does not cross over the boundary of original input pixel, thus the range of applicable information get increased. This method uses the following information; the Euclidean distance between an input pattern and modified pattern, the output of SVM, the support vector output of hidden neuron that is the closest to the initial input pattern. For the recognition experiment, 45 different combinations of USPS digit data are used. Better recognition performance is seen when selective attention is applied along with SVM than SVM only. Also, the proposed selective attention shows better performance than common selective attention.

• Analytical Science and Technology
• /
• v.17 no.4
• /
• pp.355-362
• /
• 2004

Fission products and Inter reaction layer of $U_3Si-Al$ dispersion fuel, irradiated in HANARO research reactor with 121 kW/m of maximum liner power and 63 at% of average burn-up, was characterization by EPMA (Electron Probe Micro Analyzer). The fuel punching system developed by Irradiated Materials Examination Facility (IMEF) has used to make these samples for the EPMA. With this system a very small and thin specimen which is 1.57 mm in diameter and 2 mm in thickness respectively has been fabricated to protect the EPMA operator from high radioactive fuel and to mini-mize the equivalent dose rate less than 150 mSv/h. EPMA was performed to observe layers of sectional, Inter-reaction and oxide with specimens of cutting and polished. Stoichiometry in the Inter-reaction layer with $16{\mu}m$ of thickness was $U_{2.84}$ Si $Al_{14}$ with calibration of $UO_2$ and $U_{3.24}$ Si $Al_{14.1}$ with calibration of standard specimen. metallic precipitates in this layer were not observed using fission products examination.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.22 no.2
• /
• pp.67-75
• /
• 2021

The structure of universities needs to be adjusted and reformed to cope with the decrease in admission resources and the quality of education due to the low birth rate and aging population. Such a policy is receiving much attention. To analyze the relative efficiency of private universities in Korea from the perspective of resource and performance, this study evaluated the efficiency of private university operation by applying a DEA(Data Envelopment Analysis) technique. The DEA measurements were compared with the diagnosis results of the department of education (Government) in 2018. The input and output variables used in the research analysis were utilized by the university's notification materials (public disclosure information). An analysis of the operational efficiency showed that 48% (12 universities) of the 25 DMUs (Decision Making Unit) were efficient for DEA-BCC models and that some of the capacity-building universities were operating efficiently. In addition, the DEA analysis found ways to improve inefficient groups through DEA-Additive results. This paper can be meaningful because it confirmed the relative efficiency of private universities and suggested improvement directions through the DEA method, which is characterized by the simultaneous consideration of various input and output factors. This will help apply the limited resources related to the input and output elements of each university.

• Membrane Journal
• /
• v.34 no.1
• /
• pp.70-78
• /
• 2024

A photo-crosslinked anion exchange membrane (AEM) based on quaternary-aminated polyacrylates was developed for reverse electrodialysis (RED). Although reverse electrodialysis is a clean and renewable energy generation system, the low power output and high membrane cost are serious obstacles to its commercialization. Cross-linked AEMs without any polymer supporters were fabricated through photo-crosslinking between polymer-typed acrylates with anion conducting groups, in particular, polymer-typed acrylates were synthesized based on engineering plastic with outstanding mechanical and chemical property. The fabricated membranes showed outstanding physical, chemical, and electrochemical properties. The area resistance of the fabricated membranes (CQAPPOA-20, CQAPPOA-35, and CQAPPOA-50) were ~50% lower than that of AMV (2.6 Ω cm²). Moreover, the transport number of CQAPPOA-35 wase comparable to that of AMV, despite the thin thickness (40 ㎛) of the fabricated membranes. The RED stack with the CQAPPOA-35 membrane provided an excellent maximum power density of 2.327 W m^-2 at a flow rate of 100 mL min^-1, which is 15% higher than that (2.026 W m^-2) of the RED stack with the AMV membrane. Considering easy fabrication process by UV photo-crosslinking and outstanding RED stack properties, the CQAPPOA-35 membrane is a promising candidate for REDs.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2005.11a
• /
• pp.353-356
• /
• 2005

In this paper, the proposed $gamma({\gamma})$ line system is developed for reducing the error between non-linear gamma curve produced by a formula and result produced by hardware implementation. The proposed algorithm and system is based on the specific gamma value 2.2, namely the formula is represented by {0,1}$^{2.2}$ and the bit width of input and out data is 10bit. In order to reduce the error, the system is using least squares polynomial of the numerical method which is calculating the best fitting polynomial through a set of points. The proposed gamma line is consisting of nine kinds of quadratic equations, each with their own overlap sections to get more precise. Based on the algorithm verified by $MATLAB^{TM}$ 7.0, the proposed system is implemented by using Verilog-HDL. The proposed system has 2 clock latency; 1 result per clock. The error range (LSB) is -4 and +3. Its standard deviation is 1.287956238. The total gate count of system is 2,083 gates and the maximum timing is 15.56[ns].