• Proceedings of the Korea Water Resources Association Conference
• /
• 2017.05a
• /
• pp.71-71
• /
• 2017

하천에서 평수기 유량측정은 도섭법을 이용하기 위한 지점식 측정보다는 초음파 도플러 유속계(ADCP, Acoustic Doppler Current Profiler)를 보트에 탑재하여 운용하는 측정 방식이 점차 일반화되고 있다. ADCP는 초음파의 도플러효과를 이용하여 수심이나 횡방향의 유속 분포를 측정할 수 있는 측정 장비로 일반적으로 사용되는 down-looking ADCP는 수심방향의 유속분포와 수심을 측정하여 보트의 이동속도와의 벡터 내적을 이용하여 유량을 산정하게 된다. 그러나, 이동식 ADCP 유량 측정 성과의 불확도는 제공되지 않고 있는 상황인데, 이는 불확도 산정 표준안 미비, 유속 및 수심 등 측정 요소의 관측 환경 별 불확도 정보 부족, 불확도를 산정할 수 있는 툴의 부재 등에 기인한다. 본 연구에서는 이동식 ADCP 불확도 산정 표준안을 개발하고 현장 실험을 통해 불확도 요인에 대한 규명, 불확도를 편리하게 산정할 수 있는 툴을 개발하고자 하였다. 불확도 산정 표준안으로 최근 WMO를 위시한 국제적으로 하천 유량 측정 불확도 표준안으로 채택되고 있는 GUM(Guide to the Expression of Uncertainty Measurement)을 기반으로 이동식 ADCP 유량 산정 알고리즘을 적용하여 불확도 적용 기법을 개발하였다. GUM 표준안을 기반으로 유량 측정불확도를 산정하기 위한 불확도 요인분석은 실규모 하천의 특성을 대부분 모의할 수 있는 한국건설기술연구원의 안동하천실험센터에서 수행된 실험자료를 기반으로 다양한 인자들에 대한 요소 별 불확도 분석을 수행하였다. GUM 표준안에 의하면 불확도 요인들은 오차전파의 법칙에 기반하여 전체 불확도에 전파되며, 이렇게 합성된 불확도는 t-분포의 신뢰수준 95%일 경우의 보정계수 2를 곱하여 최종적으로 확장불확도를 산정하게 된다. 이동측정방식의 ADCP의 경우 GUM 표준안에 적용하여 불확도를 평가하기 위해서 사용되는 수식이 방대하고, 매우 복잡하기 때문에 이를 실무자가 평가하기에는 한계가 있다. 이에 따라 본 연구에서는 ADCP의 유량 측정불확도를 보다 편리하게 평가하기 위하여 ADCP 유량 측정불확도 평가 소프트웨어인 AQUA(ADCP Discharge Uncertainty Assesment)를 개발하였으며, 이를 통해 실무자나 연구자들이 ADCP의 불확도 평가에 보다 편리하게 접근할 수 있을 것이라 판단된다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2018.05a
• /
• pp.383-383
• /
• 2018

최근 기후변화와 도시화로 인해 국지성 집중호우 및 불투수면적이 증가하고 있는 실정이며, 도시 지역 내의 첨두유량, 도달시간, 지체시간 등과 같은 수문학적 인자가 변화함에 따라 재산피해, 인명피해가 발생하고 있다. 저영향개발(Low Impact Development, LID) 기법은 수리수문학적 및 환경생태학적 문제를 저감하는 방안 중 하나로써 도시지역에서 수환경을 자연상태로 복원하는 대안으로 제시되고 있다. LID 기법 중 하나인 옥상녹화는 도시 내의 불투수면 증가로 인한 초과 지표면유출을 저감시켜 물관리를 하는 기술이다. 본 연구는 경남 양산시 부산대학교 제 2 캠퍼스에 조성된 옥상녹화 장치를 이용하여 정량적으로 유출량을 분석하였다. 비식생구와 식생구를 설치하고 실험의 시나리오는 강우강도를 25, 50, 75, 100 mm/hr로 설정하여 측정된 데이터 값을 바탕으로 SWMM(Storm Water Management Model) 모델링을 수행하였다. 유출량 값은 SWMM 5의 매개변수 추정지원 시스템인 SWMM-SCE를 이용하여 모형을 자동보정하였다. 보정된 모의유량은 실측유량과 0.28~3.81% 만큼의 오차를 보였고 각 시나리오에 따라 검증한 결과 상관계수가 0.82 이상으로서 실측값과 높은 상관성을 나타내었다. 옥상녹화 실험의 경우, 강우강도 75mm/hr일 때 첨두유출저감율과 지연시간은 각각 15.45% 감소, 15초 지연으로 최적의 효율이 나타났으며 강우강도 25mm/hr일 때 첨두유출저감율과 지연시간은 각각 1.36% 감소, 4초 지연으로 최저의 효율이 나타났다. SWMM 모의 결과는 강우강도 75mm/hr일 때 첨두유출저감율과 지연시간은 각각 15.45% 감소, 16초 지연으로 최적의 효율이 나타났으며 강우강도 25mm/hr일 때 첨두유출저감율과 지연시간은 각각 2.73% 감소, 4초 지연으로 최저의 효율이 나타났다.

• Journal of Korea Water Resources Association
• /
• v.54 no.spc1
• /
• pp.1023-1035
• /
• 2021

The electricity cost of a desalination facility was also predicted and reviewed, which allowed the proposed model to be incorporated into the future design of such facilities. Input data from 2003 to 2014 of the Korea Hydrographic and Oceanographic Agency (KHOA) were used, and the structure of the model was determined using the trial and error method to analyze as well as hyperparameters such as salinity and seawater temperature. The future seawater quality was estimated by optimizing the prediction model based on machine learning. Results indicated that the seawater temperature would be similar to the existing pattern, and salinity showed a gradual decrease in the maximum value from the past measurement data. Therefore, it was reviewed that the electricity cost for seawater desalination decreased by approximately 0.80% and a process configuration was determined to be necessary. This study aimed at establishing a machine-learning-based prediction model to predict future water quality changes, reviewed the impact on the scale of seawater desalination facilities, and suggested alternatives.

• Journal of the Korean Society of Radiology
• /
• v.18 no.2
• /
• pp.93-99
• /
• 2024

The digital medical imaging, especially, computed tomography (CT), should necessarily be considered in terms of noise distribution caused by converting to X-ray photon to digital imaging signal. Recently, the denoising technique based on deep learning architecture is increasingly used in the medical imaging field. Here, we evaluated noise reduction effect according to various noise types based on the U-net deep learning model in the lung CT images. The input data for deep learning was generated by applying Gaussian noise, Poisson noise, salt and pepper noise and speckle noise from the ground truth (GT) image. In particular, two types of Gaussian noise input data were applied with standard deviation values of 30 and 50. There are applied hyper-parameters, which were Adam as optimizer function, 100 as epochs, and 0.0001 as learning rate, respectively. To analyze the quantitative values, the mean square error (MSE), the peak signal to noise ratio (PSNR) and coefficient of variation (COV) were calculated. According to the results, it was confirmed that the U-net model was effective for noise reduction all of the set conditions in this study. Especially, it showed the best performance in Gaussian noise.

• Applied Chemistry for Engineering
• /
• v.28 no.5
• /
• pp.559-564
• /
• 2017

In this study, the optimization process was carried out by using the central composite model of the response surface methodology in waste cooking oil based biodiesel production process. The acid value, reaction time, reaction temperature, methanol/oil molar ratio, and catalyst amount were selected process variables. The response was evaluated by measuring the FAME content (more than 96.5%) and kinematic viscosity (1.9~5.5 cSt). Through basic experiments, the range of optimum operation variables for the central composite model, such as reaction time, reaction temperature and methanol/oil molar ratio, were set as between 45 and 60 min, between 50 and $60^{\circ}C$, and between 8 and 12, respectively. The optimum operation variables, such as biodiesel production reaction time, temperature, and methanol/oil molar ratio deduced from the central composite model were 55.2 min, $57.5^{\circ}C$, and 10, respectively. With those conditions the results deduced from modeling were as followings: the predicted FAME content of the biodiesel and the kinematic viscosity of 97.5% and 2.40 cSt, respectively. We obtained experimental results with deduced operating variables mentioned above as followings: the FAME content and kinematic viscosity of 97.7% and 2.41 cSt, respectively. Error rates for the FAME content and kinematic viscosity were 0.23 and 0.29%, respectively. Therefore, the low error rate could be obtained when the central composite model among surface reaction methods was applied to the optimized production process of waste cooking oil raw material biodiesel.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.38 no.4
• /
• pp.587-593
• /
• 2018

The current road traffic noise prediction programs of Korea, which are widely used, are based upon foreign prediction model. Thus, it is necessary to verify foreign prediction models to find out whether they are suitable for the domestic road traffic environment. In addition, an analysis and an in-depth study on the main factors should be conducted in advance as the influence factors on the occurrence of traffic noise vary for each prediction model. Therefore, this study examined the influence factors and the existing prediction models used to forecast road traffic noise. Also, analyzed their relationship with the factors influencing the noise generated by driving vehicles through multiple regression analysis using a prediction model, taking into consideration of the traffic environment and the road geometric structure. In addition, this study will apply experimental values to the existing road traffic noise prediction model (NIER, RLS-90) and the deducted road traffic noise prediction model. As a result, the order of the absolute value sum of the errors are NIER, RLS-90, model value. Through comparison and verification, developed models are to be analyzed for providing basic research results for future study on road traffic noise prediction modeling.

• Journal of the Korea Concrete Institute
• /
• v.22 no.3
• /
• pp.335-343
• /
• 2010

Steel corrosion in reinforced concrete (RC) structures is a critical problem to structural safety and many researches are being actively conducted on developing methods to maintain the required performance of the RC structures during their intended service lives. In this study, concrete mixture proportioning technique through genetic algorithm (GA) for RC structures under carbonation, which is considered to be serious in underground site and big cities, is investigated. For this, mixture proportions and diffusion coefficients of $CO_2$ from the previous researches were analyzed and fitness function for $CO_2$ diffusion coefficient was derived through regression analysis. This function based on the 12 experimental results consisted of 5 variables including water-cement ratio (W/C), cement content, sand percentage, coarse aggregate content per unit volume of concrete in unit, and relative humidity. Through genetic algorithm (GA) technique, simulated mixture proportions were proposed for 3 cases of verification and they showed reasonable results with less than relative error of 10%. Finally, assuming intended service life, different exposure conditions, design parameters, intended $CO_2$ diffusion coefficients, and cement contents were determined and related mixture proportions were simulated. This proposed technique is capable of suggesting reasonable mix proportions and can be modified based on experimental data which consider various mixing components like mineral admixtures.

• Applied Chemistry for Engineering
• /
• v.30 no.5
• /
• pp.606-614
• /
• 2019

A mixing ratio of the oil in water (O/W) emulsion of palm oil and the non-ionic surfactant (Tween-Span type) possessing different hydrophile-lipophilie balance (HLB) values was evaluated in this work. An optimum condition was determined through analysis of main and interaction effects of each quantitative factor using central composite design model-response surface methodology (CCD-RSM). Quantitative factors used by CCD-RSM were an emulsification time, emulsification speed, HLB value and amount of surfactant. On the other hand, the reaction parameters were the viscosity and mean droplet size of O/W emersion. Optimized conditions obtained from CCD-RSM were the emulsification time of 12.7 min, emulsification speed of 5,551 rpm, HLB value of 8.0 and amount of surfactant of 5.7 wt.%. Ideal experimental results under the optimized experimental condition were the viscosity of 1,551 cP and mean droplet size of 432 nm which satisfy the targeted values. The average error value from our actual experiment for verifying the conclusions was below to 2.5%. Therefore, a high favorable level could be obtained when the CCD-RSM was applied to the optimized palm oil to water emulsification.

• Radiation Oncology Journal
• /
• v.16 no.1
• /
• pp.71-79
• /
• 1998

Purpose : The objective of this study is to introduce our installation of a non-commercial 3D Planning system, Plunc and confirm it's clinical applicability in various treatment situations. Materials and Methods : We obtained source codes of Plunc, offered by University of North Carolina and installed them on a Pentium Pro 200MHz (128MB RAM, Millenium VGA) with Linux operating system. To examine accuracy of dose distributions calculated by Plunc, we input beam data of 6MV Photon of our linear accelerator(Siemens MXE 6740) including tissue-maximum ratio, scatter-maximum ratio, attenuation coefficients and shapes of wedge filters. After then, we compared values of dose distributions(Percent depth dose; PDD, dose profiles with and without wedge filters, oblique incident beam, and dose distributions under air-gap) calculated by Plunc with measured values. Results : Plunc operated in almost real time except spending about 10 seconds in full volume dose distribution and dose-volume histogram(DVH) on the PC described above. As compared with measurements for irradiations of 90-cm 550 and 10-cm depth isocenter, the PDD curves calculated by Plunc did not exceed $1\%$ of inaccuracies except buildup region. For dose profiles with and without wedge filter, the calculated ones are accurate within $2\%$ except low-dose region outside irradiations where Plunc showed $5\%$ of dose reduction. For the oblique incident beam, it showed a good agreement except low dose region below $30\%$ of isocenter dose. In the case of dose distribution under air-gap, there was $5\%$ errors of the central-axis dose. Conclusion : By comparing photon dose calculations using the Plunc with measurements, we confirmed that Plunc showed acceptable accuracies about $2-5\%$ in typical treatment situations which was comparable to commercial planning systems using correction-based a1gorithms. Plunc does not have a function for electron beam planning up to the present. However, it is possible to implement electron dose calculation modules or more accurate photon dose calculation into the Plunc system. Plunc is shown to be useful to clear many limitations of 2D planning systems in clinics where a commercial 3D planning system is not available.

• Journal of the Korean Geotechnical Society
• /
• v.38 no.7
• /
• pp.5-24
• /
• 2022

The current performance evaluation of slope anchors qualitatively determines the physical bonding between the anchor head and ground as well as cracks or breakage of the anchor head. However, such performance evaluation does not measure these primary factors quantitatively. Therefore, the time-dependent management of the anchors is almost impossible. This study is an evaluation of the 3D numerical model by SfM which combines UAS images with terrestrial LiDAR to collect numerical data on the damage factors. It also utilizes the data for the quantitative maintenance of the anchor system once it is installed on slopes. The UAS 3D model, which often shows relatively low precision in the z-coordinate for vertical objects such as slopes, is combined with terrestrial LiDAR scan data to improve the accuracy of the z-coordinate measurement. After validating the system, a field test is conducted with ten anchors installed on a slope with arbitrarily damaged heads. The damages (such as cracks, breakages, and rotational displacements) are detected and numerically evaluated through the orthogonal projection of the measurement system. The results show that the introduced system at the resolution of 8K can detect cracks less than 0.3 mm in any aperture with an error range of 0.05 mm. Also, the system can successfully detect the volume of the damaged part, showing that the maximum damage area of the anchor head was within 3% of the original design guideline. Originally, the ground adhesion to the anchor head, where the z-coordinate is highly relevant, was almost impossible to measure with the UAS 3D numerical model alone because of its blind spots. However, by applying the combined system, elevation differences between the anchor bottom and the irregular ground surface was identified so that the average value at 20 various locations was calculated for the ground adhesion. Additionally, rotation angle and displacement of the anchor head less than 1" were detected. From the observations, the validity of the 3D numerical model can obtain quantitative data on anchor damage. Such data collection can potentially create a database that could be used as a fundamental resource for quantitative anchor damage evaluation in the future.