• Proceedings of the Korean Society for Agricultural Machinery Conference
• /
• 2017.04a
• /
• pp.166-166
• /
• 2017

모돈의 건강 상태를 정량 지수화 하기 위한 연구를 수행 중이다. 지제이상, 섭식 불량, 수면 패턴 등의 운동 특성 분석을 위하여 복수의 초음파 센서를 이용하였다. 시계열 계측 신호를 분석하여 정량 지수화를 수행하는 과정에서 주파수 도메인 분석을 시도하였다. 이 과정에서 주파수 도메인의 분해능에 따른 편차 극복을 위한 비선형 모델링을 수행하였다. 또한 인접한 시계열 데이터 구간 간의 상관성 분석이 가능하면 대용량 데이터의 실시간 처리로 인한 지연 시간 극복 및 기대되는 예후에 대한 조기 진단이 가능할 것이다. 본 연구에서는 구글에서 제공하는 Tensorflow와 NVIDIA에서 제공하는 CUDA 엔진을 동시 적용한 심층 학습 시스템을 이용하였다. 전 처리를 위하여 주파수 분해능 (2분, 3분, 5분, 7분, 11분, 13분, 17분, 19분)에 따른 데이터 집합을 1단계로 두고, 상위 10 순위 안에 드는 파워 스펙트럼 밀도의 크기를 2단계로 하여, 총 2~10개의 입력 노드를 순차적으로 선정하였고, 동일한 방식으로 인접한 시계열의 파워 스펙터럼 밀도를 순위를 변화시켜 지정하였다. 대표적인 심층학습 모델인 Softmax regression with a multilayer convolutional network를 이용하여 Recursive feature selection 경우의 수를 $8{\times}9{\times}9$로 총 648 가지 선정하고, Epoch는 10,000회로 지정하였다. Calibration 모델링의 경우 Cost function이 10% 이하인 경우 해당 경우의 학습을 중단하였으며, 모델 간 상호 교차 검증을 수행하기 위하여 $_8C_2{\times}_8C_2{\times}_8C_2$ 경우의 수에 대한 Verification test를 수행하였다. Calibration 과정 상 모든 경우에 대하여 10% 이하의 Cost function 값을 보였으나, 검증 테스트 과정에서 모든 경우에 대하여 $r^2$ < 0.5 인 결정 계수 값이 나타났다. 단적으로 심층학습 모델의 과도한 적합(Over fitting) 방식의 한계를 보인 것이라 판단할 수 있다. 적합한 Feature selection 및 심층 학습 모델에 대한 지속적이고 추가적인 고려를 통해 과도적합을 해소함과 동시에 실효적이고 활용 가능한 Classification을 위한 입, 출력 노드 단의 전후 Indexing, Quantization에 대한 고려가 필요할 것이다. 이를 통해 모돈 생체 정보 정량화를 위한 지능형 현장 진단 기술 연구를 지속할 것이다.

• Applied Chemistry for Engineering
• /
• v.9 no.3
• /
• pp.383-387
• /
• 1998

The effect of hydrophilic surface modification of low density polyethylene(LDPE) byt the plasma gas($O_2$, $N_2$, and $O_2/N_2$) was investigated from the point of view of the functionalities of the generated LDPE surfaces and the contact angle. By virtue of x-ray photoelectron spectra(XPS) and attenuated total reflectance(FT-IR ATR) analysis, the LDPE surfaces treated with plasma were generated with oxygen functionalities of carbonyl, carboxyl, and the like, nitrogen functionalities by nitrogen plasma and mixing of nitrogen and oxygen plasma treatment were identified with. It was found that nitrogen plasma treatment showed with minimum value at contact angle for rf-power and treatment time, we had obtained optimum condition for hydrophilic surface modification at composite parameter, [(W/FM)t] 520~550GJs/kg.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.23 no.3
• /
• pp.214-219
• /
• 2013

This paper proposes the robust high-gain observer based SOC estimatro for uncertain RC model of Li-Ion batteries. In general, RC battery model has inevitable uncertainties and it cause some negative effect to estimate the accurate SOC of Li-Ion batteries. The proposed estimator overcomes such weakness with two techniques; high-gain observer design technique and sliding mode control technique. A high-gain observer provides the robustness against model uncertainties to the proposed estimator. A sliding mode control technique helps the proposed estimator by reducing the side effect of adopting a high-gain observer such as peaking phenomenon and perturbation. The performance of the proposed estimator is verified by some simulation.

• Journal of Sensor Science and Technology
• /
• v.3 no.2
• /
• pp.65-73
• /
• 1994

TO:F($SnO_{2}:F$) thin films were prepared by RF magnetron sputtering system. The dependence of their structural, electrical, and optical properties on deposition conditions such as substrate temperature, working pressure and power was studied. The optimum conditions of TO:F thin film are $SnF_{2}$ content of 15wt.% in target, RF power of 150W, substrate temperature of $150^{\circ}C$ and working pressure of 2mmTr. The resistivity and transmittance at 550nm in visible spectrum of the TO:F film deposited at optimum condition are $9{\times}10^{-4}{\Omega}{\cdot}cm$ and above 85%, respectively. For the films deposited from the target without $SnF_{2}$ and with 15wt.% $SnF_{2}$, the optical bandgaps calculated from the transmittance curves are 3.84 and 3.9eV, respectively. X-ray diffraction patterns showed that TO and TO:F films had tetragonal rutile structure with (101), (200) direction.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.30 no.4
• /
• pp.106-113
• /
• 2002

In spacecraft system, structure subsystem has the mission of supporting all the components safely under various space environmental conditions. The safety of spacecraft structure is finally verified from the coupled load analysis, which is a branch of load analysis which combines the launch vehicle and satellite. This study introduces the optimization algorithm to reduce the weight of spacecraft structure under launch environmental conditions directly. The acceleration responses are obtained by the introduction of coupled load analysis, which lead to check the failure of spacecraft structural members. The results show a 12% saving of structural weight and this saving is mainly driven by the thickness of honeycomb core, which strongly affects the natural frequencies of platforms and panels.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.11 no.12
• /
• pp.5120-5126
• /
• 2010

ABLB(Alternate Binaural Loudness Balance) test is one of the medical assessments to diagnose detailed lesion of sensory-neural hearing loss based on a recruitment phenomenon. This paper describes an Auto ABLB test software. This software automatically decides test frequency and intensity of first test tone. And it supports an automatic algorithm that analyses the subject's reponses in the current step and then decides intensity of tone provided in next step. Also, this software supports an interface to control and monitor subject's ABLB test. The assessment result is represented by a ladder diagram. The output of this software has been verified using a digital oscilloscope.

• 한국HCI학회:학술대회논문집
• /
• 2009.02a
• /
• pp.696-700
• /
• 2009

In this paper, we present a framework which enables users to interact naturally with hand gestures on a digital table. In general tabletop applications, one gesture is mapped to one function or command. Therefore, users should know these relations, and make predefined gestures as input. In contrast, users can make input gesture without cognitive load in our system. Instead of burdening users, the system possesses knowledge about gesture interaction, and infers proactively users' gestures and intentions. When users make a gesture on the digital surface, the system begins to analyze the gestures and designs the response according to users' intention.

• The Journal of the Acoustical Society of Korea
• /
• v.13 no.4
• /
• pp.40-49
• /
• 1994

In this paper, we propose methods for improving the performance of word recognition system. The ray stratey of the first method is to apply the inertia to the feature vector sequences of speech signal to stabilize the transitions between VQ cdoes. The second method is generating the new observation probabilities using the transition matrix of VQ codes as weights at the observation probability of the output symbol, so as to take into account the time relation between neighboring frames in DHMM. By applying the inertia to the feature vector sequences, we can reduce the overlapping of probability distribution of the response paths for each word and stabilize state transitions in the HMM. By using the transition matrix of VQ codes as weights in conventional DHMM. we can divide the probability distribution of feature vectors more and more, and restrict the feature distribution to a suitable region so that the performance of recognition system can improve. To evaluate the performance of the proposed methods, we carried out experiments for 50 DDD area names. As a result, the proposed methods improved the recognition rate by $4.2\%$ in the speaker-dependent test and $12.45\%$ in the speaker-independent test, respectively, compared with the conventional DHMM.

• Journal of Internet Computing and Services
• /
• v.11 no.5
• /
• pp.1-8
• /
• 2010

This paper presentsthe performance study of blind equalizer algorithms for impulsive-noise environments based on Gaussian kernel and constant modulus error(CME). Constant modulus algorithm(CMA) based on CME and mean squared error(MSE) criterion fails in impulsive noise environment. Correntropy blind method recently introduced for impulsive-noise resistance has shown in PAM system not very satisfying results. It is revealed in theoretical and simulation analysis that the maximization of zero-error probability based on CME(MZEP-CME) originally proposed for Gaussian noise environments produces superior performance in impulsive noise channels as well. Gaussian kernel of MZEP-CME has a strong effect of becoming insensitive to the large differences between the power of impulse-infected outputs and the constant modulus value.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.11 no.2
• /
• pp.696-702
• /
• 2010

In many industrial applications, it is very important to control the temperature of the controlled object to the target temperature as closely as possible. Although it is apparent that the great obstacles in controller design are time-delay of the thermal responses of the controlled object and the effect of thermal interference between neighboring heating zones, one more fundamental obstacle is a very large amount of time which is required during repeated experiments in controller design process. Therefore, a convenient simulation environment, which can represent thermal behavior accurately within appropriate time, is needed. In this paper, a prototype of 2D FEM (finite element method) heat transfer simulation environment using MATLAB is constructed to be usefully adopted into industrial applications with temperature controller design.