• 대한후두음성언어의학회지
• /
• 제13권1호
• /
• pp.28-32
• /
• 2002

Background and Objective : Classical singing students who have received in systematic voice training appeared distinctive voice characteristics compared to normal people who have not received in systematic voice training. The purpose of this study was to determine the correlation between acoustic parameters and Electroglottography(EGG) parameters in two groups(ordinary college students vs. classical singing students group). Materials and Methods : The 80 ordinary college students and 65 classical singing students participated in this study by utilizing Dr. speech program to obtain acoustic measurements and physiologic measurements simultaneously. The Pearson correlation coefficient was used to find the correlation between acoustic parameters and EGG parameters in two groups(ordinary college students group and classical singing students group). Results : The results of the study were as follows : First, there was no correlation between Jitter and EGG Jitter in ordinary college students group, but there was strong correlation between Jitter and EGG Jitter in classical singing students group. Second, there was no correlation between Shimmer and EGG Shimmer in ordinary college students group, but there was strong correlation between Shimmer and EGG Shimmer in classical singing students group. Third, there was no correlation between Harmonic to Noise Ratio(HNR) and EGG HNR in ordinary college students group, but there was strong correlation between HNR and EGG HNR in classical singing students group. Finally, there was no correlation between Normalized Noise Energy(NNE) and EGG NNE in two groups.

• Smart Structures and Systems
• /
• 제30권6호
• /
• pp.673-686
• /
• 2022

Deep learning algorithms for Structural Health Monitoring (SHM) have been extracting the interest of researchers and engineers. These algorithms commonly used loss functions and evaluation indices like the mean square error (MSE) which were not originally designed for SHM problems. An updated loss function which was specifically constructed for deep-learning-based structural damage detection problems has been proposed in this study. By tuning the coefficients of the loss function, the weights for damage localization and quantification can be adapted to the real situation and the deep learning network can avoid unnecessary iterations on damage localization and focus on the damage severity identification. To prove efficiency of the proposed method, structural damage detection using convolutional neural networks (CNNs) was conducted on a truss bridge model. Results showed that the validation curve with the updated loss function converged faster than the traditional MSE. Data augmentation was conducted to improve the anti-noise ability of the proposed method. For reducing the training time, the normalized modal strain energy change (NMSEC) was extracted, and the principal component analysis (PCA) was adopted for dimension reduction. The results showed that the training time was reduced by 90% and the damage identification accuracy could also have a slight increase. Furthermore, the effect of different modes and elements on the training dataset was also analyzed. The proposed method could greatly improve the performance for structural damage detection on both the training time and detection accuracy.

• 대한기계학회논문집A
• /
• 제28권6호
• /
• pp.816-825
• /
• 2004

In this work, effects of hyper-elastic rubber material properties on the indentation load-deflection curve and subindenter deformation are examined via finite element (FE) analyses. An optimal location for data analysis is selected, which features maximum strain energy density and negligible frictional effect. We then contrive two normalized functions, which map an indentation load vs. deflection curve into a strain energy density vs. first invariant curve. From the strain energy density vs. first invariant curve, we can extract the rubber material properties. This new spherical indentation approach produces the rubber material properties in a manner more effective than the common uniaxial tensile/com-pression tests. The indentation approach successfully measures the rubber material properties and the corresponding nominal stress-strain curve with an average error less than 3%.

• 한국콘텐츠학회:학술대회논문집
• /
• 한국콘텐츠학회 2004년도 추계 종합학술대회 논문집
• /
• pp.495-500
• /
• 2004

The neutron capture cross section of $^{99}Tc$ has been measured relative to the $^{10}B$(n,g) standard cross section by the neutron time-of-flight(TOF) method in the energy range of 0.007 eV to 47keV using a 46-MeV electron linear accelerator(linac) at the Research Reactor Institute, Kyoto University(KURRI). In order to experimentally prove the result obtained, the supplementary cross section measurement has been made from 0.3 eV to 1 keV using the Kyoto University Lead slowing-down Spectrometer(KULS) coupling to the linac. The relative measurement by the TOF method has been normalized to the reference value(20.01 b) at 0.0253 eV and the KULS measurement to that by the TOF method. The existing experimental data and the evaluated capture cross sections in ENDF/B-VI, JENDL-3.2, and JEF-2.2 have been compared with the current measurements by the linac TOF and the KULS experiments. The energy dependency of the KULS data is close to that of the TOF data which are energy-broadened by the resolution function of the KULS.

• 한국음향학회지
• /
• 제16권3호
• /
• pp.18-25
• /
• 1997

본 연구에서는 성문파에 의한 음원 특성과 입술에서 음성이 방사될 때 발생하는 방사 특성을 효과적으로 제거하기 위하여 성문파의 음원 특성과 입술에서의 방사 특성은 근사적으로 음파의 에너지와 비례한다고 가정하고, 정규화된 단구간 에너지를 이용하는 에너지 의존 프리엠퍼시스 기법을 제안하며, 이를 이용하여 비안정 구간인 발음의 시작 부분과 천이구간에 대하여 음성 신호 분석 성능을 개선하고자 한다. 제안된 프리엠퍼시스 기법을 이용하여 5개 한국어 단모음의 스펙트럼 및 형성음 주파수 추출 등의 음성 신호 분석을 수행하고, 기존에 널리 이용되던 두 가지 프리엠퍼시스 기법과 성능을 비교하여 본 결과, 제안된 방법에 의한 스펙트럼의 형태가 기존의 방법에 비하여 상당히 개선되었고, 보다 더 정확한 형성음 주파수를 나타내며, 인접한 두 형성음 주파수가 증첩되는 현상이 제거되었음을 알 수 있었다.

• Elastomers and Composites
• /
• 제39권1호
• /
• pp.23-35
• /
• 2004

본 연구에서는 먼저 유한요소해석을 통해 주요 물성계수들이 압입시 하중-변위 곡선형상에 미치는 영향을 분석하였다. 또한 유한요소 압입해석을 통해 마찰계수의 영향으로 하중-변위 곡선, 시편하부의 단위부피당 변형에너지 및 변형률 주불변량이 바뀌지 않는 최적 압입깊이와 시편하부지점을 선정하였다. 이러한 관찰을 통해 하나의 요소에서 얻어지는 단위부피당 변형 에너지와 변형률 주불변량을 하중-변위 데이터와 모사 시킬 수 있는 무차원 함수를 얻을 수 있었으며, 이 과정에서 예측된 물성계수를 바탕으로 공칭응력-공칭변형률 곡선을 얻을 수 있었다.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2003년도 Proceedings of ACRS 2003 ISRS
• /
• pp.351-353
• /
• 2003

We present a retrieval scheme for the remote sensing of evapotranspiration (ET) over rice paddy. To perform the retrieval, high-resolution airborne imagery of multi-spectral visible and thermal infrared data, and ground-based meteorological measurements are utilized. Our ET retrieval scheme is based on the basic principal of surface energy budget, which is a result of balance in longwave and shortwave radiation, latent heat, sensible heat, and energy flux into the ground. To partition the latent and sensible heat fluxes of interest from the energy balance equation, three basic parameters are of most concern, including albedo, surface temperature, and normalized difference vegetation index (NDVI). The NDVI and albedo can be easily derived from the visible and near infrared spectral data, while the surface tem-perature can be determined through the analysis of the infrared data with the Stefan Boltzmann law. From the airborne imagery taken on 28 April 2003, we observe very good dry and wet pixels that can be easily corre-sponded to the radiation and evaporation controlled crite-ria, respectively, and, hence, for the further use in defin-ing the evaporative fraction needed to partition sensible and latent heat fluxes from the net energy flux. The de-rived ET is compared with the in situ measurements.

• Journal of Advanced Marine Engineering and Technology
• /
• 제22권3호
• /
• pp.343-352
• /
• 1998

Higher frequency of energy transfer or at least energy conversion has to be used in order to reduce the size of inductors and capacitors required in the power supplies. Conventional PWM switching-mode power supplies have a limitation of operating frequency due to switching losses in the switching transistors and rectifier diodes. Means of reducing switching losses have been developed for high-frequency resonant amplifiers or more exactly dc/ac inverters. Because of smooth current and voltage waveforms resonant convertesrs havelower device switching losses and stresses lower electromagnetic interference(EMI) and lower noise than PWM converters. Therefore in this paper design equations of Classs E resonant low dv/dt rectifier with a series resonant capacitor drived using Fourier series techniques. The theory is compared with simulation results obtained for the rectifier operating at 10[MHz] ac input and 5[V] coutput.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2006년도 정기 학술대회 논문집
• /
• pp.15-20
• /
• 2006

Real time health monitoring system was studied to detect the . generation of defects in the composite structures during service life. The PZT sensors were embedded into the woven-glass/phenol composite plate during the fabrication. VARTM (Vacuum Assisted Resin Transfer Molding) process were used to fabricate the composite plate. A Teflon tape was embedded between glass fiber layers to mimic delamination induced during service. Normalized maximum amplitude and energy analyses were used for the acquired signals. Both amplitude and energy of acquired signals were extremely sensitive to the delamination. Therefore, it was successful to detect and to locate the defects in composite plate by monitoring signals from sensors and using the proposed method.

• Journal of Advanced Marine Engineering and Technology
• /
• 제22권5호
• /
• pp.617-625
• /
• 1998

A passive energy recovery snubber for high-power-factor boost rectifier, in which the main switch is implemented with a MOSFET, is described in terms of the equivalent circuits that are operational during turn-on and turn-off sequences. These equivalent circuits are analyzed so that the overshoot voltage across the main switch, the snubber current, and the turn-off transition time can be predicted analytically. From these results, the normalized overshoot voltage is reduced to 1 as $_W2T_on$ varies from zero to $\pi$/2, and then it is fIxed at 1 for $_W2T_on$> $\pi$/2. The peak snubber inductor current is directly proportional to the input current. The turn-offtransition time wltoffvaries from 0 to 2.57, depending on $_W2T_on$. The main switch combined with proposed snubber can be turned on with zero current and turned off at limited voltage stress. The high-power-factor boost rectifier with proposed snubber is implemented, and the experimental results are presented to confirm the validity of proposed snubber.