• Fire Science and Engineering
• /
• v.26 no.5
• /
• pp.21-27
• /
• 2012

A real A real scale fire test was performed in accordance with KS F ISO 9705 test method to investigate the combustion characteristics of glass wool sandwich panels. To do this, six kinds of specimens having different density and thickness were examined. The glass sandwich panels were installed inside the room, which had internal dimensions of 2.4 m wide${\times}3.6m$ deep${\times}2.4m$ high. also, combustion characteristic are determined through the exposure of specimens to flame by the propane gas burner has a capacity of 100 kW (10 minutes) and 300 kW (10 minutes) for total 25 minutes of test time. Results of the real sale fire test, it was found that maximum HRR of each specimen was 333.2~365.5 kW, maximum heat flux was 12.4~12.9 kW/$m^2$ And, maximum internal temperature for all specimens was not over $500^{\circ}C$. During the real scale fire test, flash-over didn't occur and the difference by density and thickness of specimen was not found from the results of HRR, heat flux, and internal temperature measurement.

• Journal of IKEEE
• /
• v.27 no.4
• /
• pp.486-493
• /
• 2023

In this paper, we propose an automatic hand gesture area extraction and recognition technique using FMCW radar-based point cloud and LSTM. The proposed technique has the following originality compared to existing methods. First, unlike methods that use 2D images as input vectors such as existing range-dopplers, point cloud input vectors in the form of time series are intuitive input data that can recognize movement over time that occurs in front of the radar in the form of a coordinate system. Second, because the size of the input vector is small, the deep learning model used for recognition can also be designed lightly. The implementation process of the proposed technique is as follows. Using the distance, speed, and angle information measured by the FMCW radar, a point cloud containing x, y, z coordinate format and Doppler velocity information is utilized. For the gesture area, the hand gesture area is automatically extracted by identifying the start and end points of the gesture using the Doppler point obtained through speed information. The point cloud in the form of a time series corresponding to the viewpoint of the extracted gesture area is ultimately used for learning and recognition of the LSTM deep learning model used in this paper. To evaluate the objective reliability of the proposed technique, an experiment calculating MAE with other deep learning models and an experiment calculating recognition rate with existing techniques were performed and compared. As a result of the experiment, the MAE value of the time series point cloud input vector + LSTM deep learning model was calculated to be 0.262 and the recognition rate was 97.5%. The lower the MAE and the higher the recognition rate, the better the results, proving the efficiency of the technique proposed in this paper.

• International Journal of Highway Engineering
• /
• v.9 no.3
• /
• pp.1-12
• /
• 2007

Roads in Korea can be classified into four types according to their responsible authorities. For example, Motorway is constructed, managed, and operated by the Korea Highway Corporation. Ministry of Construction and Transportation is in charge of National Highway, and Province Roads are run by each province government. Urban/county Roads are run by corresponding local government. This study analyses the trends of road accidents for each road type. For this purpose, the numbers of accidents, fatalities, and injuries are compared for each road type for last 15 years. The result shows that Urban/County Roads are the most dangerous, while Motorways are the safest, when we simply compare the numbers of accidents, fatalities, and injuries. However, when we compare these numbers by dividing by total road length, National Highway becomes the most dangerous while Province Roads becomes the safest. In the case of road accidents, fatalities, and injuries per vehicle km, which is known as the most objective comparison measure, it turns out that National Highway is the most dangerous roads again. This study also developed time series models to estimate trends of fatalities for each road type. These models will be useful when we set up or evaluate targets of national road safety.

• Proceedings of the Acoustical Society of Korea Conference
• /
• spring
• /
• pp.82-85
• /
• 1999

본 연구에서는 연속음성인식 시스템의 성능 향상을 위하여 음성의 채록환경 및 데이터량 등을 고려한 효과적인 언어모델 작성방법을 제안하고, 이를 항공편 예약시스템에 적용하여 성능 평가 실험을 실시한 결과 $91.6\%$의 인식률을 얻어 제안한 방법의 유효성을 확인하였다. 이를 위하여 소량의 200문장의 항공편 예약 텍스트 데이터를 이용하여 좀더 강건한 단어발생 확률을 가지도록 하기 위해 일반적으로 대어휘 연속음성인식에서 많이 이용되고 있는 단어 N-gram 언어모델을 도입하고 이를 다양한 발성환경을 고려하여 1,154문장으로 확장한 후 동일 문장'을 반복 학습하여 언어모델을 작성하였다. 인식에 있어서는 오인식과 문법적 오류를 최소화하기 위하여 forward - backward pass 방법의 stack decoding알고리즘을 이용하였다. 인식실험 결과, 평가용 3인의 200문장을 각 반복학습 회수에 따라 학습한 각 언어모델에 대해 평가한 결과, forward pass의 경우 평균 $84.1\%$, backward pass의 경우 평균 $91.6\%$의 문장 인식률을 얻었다. 또한, 반복학습 회수가 증가함에 따라 backward pass의 인시률의 변화는 없었으나, forward pass의 경우, 인식률이 반복회수에 따라 증가하다가 일정값에 수렴함을 알 수 있었고, 언어모델의 복잡도에서도 반복회수가 증가함에 따라 서서히 줄어들며 수렴함을 알 수 있었다. 이상의 결과로부터 소량의 텍스트 데이터를 이용한 제한된 태스크에서 언어모델을 작성할 때 반복학습 방법이 유효함을 확인할 수 있다.

• Environmental and Resource Economics Review
• /
• v.25 no.4
• /
• pp.471-494
• /
• 2016

This study investigates the effect of remittance and agricultural aid inflows on GDP growth rate volatility in response to climate change shocks in twenty-eight African countries by using system generalized method of moments from 1996 to 2013 with three years grouped data. The climate change shocks are indicated by four variables; natural disasters, rainfall variability, fluctuation in temperature and the weighted anomaly standardized precipitation (WASP) index. Consequently, natural disasters and temperature variability have a significant effect on GDP volatility, while rainfall variability and WASP index have no adverse consequence on stabilization of the economy. On the other hand, in general, remittances and agricultural aid are helpful to stabilize the economy and especially remittances inflows can play a crucial role as insurance when natural disasters occur.

• Journal of the Korean Institute of Gas
• /
• v.9 no.3 s.28
• /
• pp.27-31
• /
• 2005

The membrane to be applied inside of the LNG storage tank is provided with corrugations to absorb thermal contraction and expansion caused by LNG temperature and pressure changes. It is very important to measure their thermal strains under LNG temperature by analytical and experimental stress analysis of the membrane. We have developed a stress measurement system using strain gages and measured the strain during cooldown and storing the LNG. We also analyzed the measured data by comparison with the FEM data.

• Proceedings of the Korean Society of Propulsion Engineers Conference
• /
• 2006.05a
• /
• pp.151-155
• /
• 2006

Elastic-viscoplastic structural analysis is performed for regenerative cooling chamber of liquid rocket thrust chamber using Bodner-Partom visco-plastic model. Strain rate test was also conducted for a copper alloy at various temperatures in order to get material constants of visco-plastic model used in the structural analysis. Material constants of visco-plastic model were obtained from strain rate test results and visco-plsstic model was incorporated into finite element program, Marc, by means of user subroutine. The structural analysis results indicate that the deformation of cooling channel is mostly caused by thermal loading rather than pressure loading and confirmed structural stability of the cooling channel under operating condition.

• Journal of Energy Engineering
• /
• v.23 no.2
• /
• pp.42-48
• /
• 2014

Soot and tar which were derived from combustion or pyrolysis processes in Puverized Coal(PC) furnace or boiler have been significantly dealing in a radiative heat transfer and an additional source of NOx. Furthermore, the increasing for the use of a coal with low caloric value gives rise to a lot of tar-soot yield and LOI in a recycled ash for using cement materials. So, the ash with higher tar-soot yield and LOI can not recycle due to decreased strength of concrete. In this study, tar-soot yields and flame structures were investigated using the LFR for a blending combustion with bituminous coal and sub-bituminous coal. Also, The investigation were conducted as each single coals and blending ratio. The coals are used in a doestic power plant. In the experimental results, sub-bituminous coal with high volatile contents shows longer soot cloud length than bituminous coal, but overall flame length was shorter than bituminous coal. Tar-soot yields of sub-bituminous coal is lower than those of bituminous coal. Combustion characteristics are different between single coal and blended coal. Therefore, finding an optimal coal blending ratio according to coal rank effects on tar-soot yields.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.27 no.5
• /
• pp.705-712
• /
• 2003

Analysis for the thermal stress distribution in the laminated plates containing a hot-spot(local heating region) is performed. It is assumed that the local heating region induces only mechanical stress by the thermal expansion but effect of the thermal conduction is neglected. The region is regarded equivalent to a homogeneous inclusion expanding in a laminated medium. As an example, Au/YBCO/Al$_2$O$_3$laminate which is often employed for High Temperature Superconducting Fault Current Limiter(HTS FCL) has been analyzed. Effects of heat input, thickness of each layer and the got spot size upon the stress distribution in the hot-spot have been investigated. For a constant heat generation into the hot-spot, as the thickness of the Al$_2$O$_3$substrate increases, the stress in the YBCO layer is peculiarly oscillated, and the curvature of laminate has a maximum at a certain thickness of the Al$_2$O$_3$.

• 한국어정보학회:학술대회논문집
• /
• 2016.10a
• /
• pp.243-245
• /
• 2016

광학 문자 인식(OCR)을 통해 문서의 글자를 인식할 때 띄어쓰기 오류가 발생한다. 본 논문에서는 이를 해결하기 위해 OCR의 후처리 과정으로 동적 프로그래밍을 이용한 분절(Segmentation) 방식의 띄어쓰기 오류 교정 시스템을 제안한다. 제안하는 시스템의 띄어쓰기 오류 교정 과정은 다음과 같다. 첫째, 띄어쓰기 오류가 있다고 분류된 어절 내의 공백을 모두 제거한다. 둘째, 공백이 제거된 문자열을 동적 프로그래밍을 이용한 분절로 입력 문자열에 대하여 가능한 모든 띄어쓰기 후보들을 찾는다. 셋째, 뉴스 기사 말뭉치와 그 말뭉치에 기반을 둔 띄어쓰기 확률 모델을 참조하여 각 후보의 띄어쓰기 확률을 계산한다. 마지막으로 띄어쓰기 후보들 중 확률이 가장 높은 후보를 교정 결과로 제시한다. 본 논문에서 제안하는 시스템을 이용하여 OCR의 띄어쓰기 오류를 해결할 수 있었다. 향후 띄어쓰기 오류 교정에 필요한 언어 규칙 등을 시스템에 추가한 띄어쓰기 교정시스템을 통하여 OCR의 최종적인 인식률을 향상에 대해 연구할 예정이다.