• 한국지하수토양환경학회:학술대회논문집
• /
• 한국지하수토양환경학회 2003년도 추계학술발표회
• /
• pp.11-24
• /
• 2003

This paper presents the results of laboratory investigation performed to study the role of different air sparging system parameters on the removal of benzene from saturated soils and groundwater. A series of one-dimensional experiments was conducted with predetermined contaminant concentrations and predetermined injected airflow rates and pressures to investigate the effect of soil type and the use of pulsed air injection on air sparging removal efficiency. On the basis of these studies, two-dimensional air sparging remediation systems were investigated to determine the effect of soil heterogeneity on the removal of benzene from three different homogeneous and heterogeneous soil profiles. This study demonstrated that the grain size of the soils affects the air sparging removal efficiency. Additionally, it was observed that pulsed air injection did not offer any appreciable enhancement to contaminant removal for the coarse sand; however, substantial reduction in system operating time was observed for fine sand. The 2-D experiments showed that air injected in coarse sand profiles traveled in channels within a parabolic zone. In well-graded sand the zone of influence was found to be wider due to high permeability and increased tortuosity of this soil type. The influence zone of heterogeneous soil (well-graded sand between coarse sand) showed the hybrid airflow patterns of the individual soil test. Overall, the mechanism of contaminant removal using air sparging from different soil conditions have been determined and discussed.

• 한국가시화정보학회지
• /
• 제19권3호
• /
• pp.92-98
• /
• 2021

Large Eddy Simulation (LES) has been popularly applied and used in the last several decades to simulate turbulent boundary layer in the numerical domain. A fully developed turbulent boundary layer has also been applied to predict the complicated wake flow behind bluff bodies. In this study we aimed to generate an artificial turbulent boundary layer, which is based on an exponential correlation function, and generates a series of realistic three-dimensional velocity data in two-dimensional inlet section which are correlated both in space and in time. The results suggest its excellent capability for high Reynolds number flows. To make an effective generation, a hexahedral mesh has been used and Cholesky decomposition was applied to possess suitable turbulent statistics such as the randomness and correlation of turbulent flow. As a result, the flow characteristics in the domain and fluctuating pressure near the wall are very close to those of fully developed turbulent boundary layers.

• 국제학술발표논문집
• /
• The 6th International Conference on Construction Engineering and Project Management
• /
• pp.559-563
• /
• 2015

Identification of energy use patterns in buildings has a great opportunity for energy saving. To find what energy use patterns exist, clustering analysis has been commonly used such as K-means and hierarchical clustering method. In case of high dimensional data such as energy use time-series, data reduction should be considered to avoid the curse of dimensionality. Principle Component Analysis, Autocorrelation Function, Discrete Fourier Transform and Discrete Wavelet Transform have been widely used to map the original data into the lower dimensional spaces. However, there still remains an ongoing issue since the performance of clustering analysis is dependent on data type, purpose and application. Therefore, we need to understand which data reduction techniques are suitable for energy use management. This research aims find the best clustering method using energy use data obtained from Seoul National University campus. The results of this research show that most experiments with data reduction techniques have a better performance. Also, the results obtained helps facility managers optimally control energy systems such as HVAC to reduce energy use in buildings.

• 한국항해학회지
• /
• 제25권4호
• /
• pp.371-379
• /
• 2001

We have reported the new measurement system which was substituted digital filter for the analog filter in order to develop the optimal system that could find the time delay between each sensors with high accuracy. And also we have confirmed through the experiments that the accuracy of measurements were differentiated by the methods what kind of digital filter had been adopted. This paper suggests two algorithms which approximate the sound source's bearing and distance. One is that sound source's relative bearing can be approximately regarded as the gradient of hyperbolic asymptote, the other is that the source's range can be approximated under the condition of a long range source relative to the sensor's interval. And a series of experiments were carried out with the source's distance 22.42meters and the random bearing interval within the limits of $-90^{\circ}$~$+90^{\circ}$. As a result, we have recognized that the approximation methods could measure the bearing and distance with higher accuracy than the method using trigonometric relation could do.

• International Journal of Internet, Broadcasting and Communication
• /
• 제13권2호
• /
• pp.208-217
• /
• 2021

3D-CNN is one of the deep learning techniques for learning time series data. However, these three-dimensional learning can generate many parameters, requiring high performance or having a significant impact on learning speed. We will use these 3D-CNNs to learn hand gesture and find the parameters that showed the highest accuracy, and then analyze how the accuracy of 3D-CNN varies through input data changes without any structural changes in 3D-CNN. First, choose the interval of the input data. This adjusts the ratio of the stop interval to the gesture interval. Secondly, the corresponding interframe mean value is obtained by measuring and normalizing the similarity of images through interclass 2D cross correlation analysis. This experiment demonstrates that changes in input data affect learning accuracy without structural changes in 3D-CNN. In this paper, we proposed two methods for changing input data. Experimental results show that input data can affect the accuracy of the model.

• Communications for Statistical Applications and Methods
• /
• 제29권1호
• /
• pp.53-64
• /
• 2022

High dimensional time series is gaining considerable attention in recent years. The sparse vector heterogeneous autoregressive (VHAR) model proposed by Baek and Park (2020) uses adaptive lasso and debiasing procedure in estimation, and showed superb forecasting performance in realized volatilities. This paper extends the sparse VHAR model by considering non-convex penalties such as SCAD and MCP for possible bias reduction from their penalty design. Finite sample performances of three estimation methods are compared through Monte Carlo simulation. Our study shows first that taking into cross-sectional correlations reduces bias. Second, nonconvex penalties performs better when the sample size is small. On the other hand, the adaptive lasso with debiasing performs well as sample size increases. Also, empirical analysis based on 20 multinational realized volatilities is provided.

• Structural Engineering and Mechanics
• /
• 제84권3호
• /
• pp.323-335
• /
• 2022

The main idea of the framework is to seamlessly combine a reasonably accurate and fast surrogate model with the importance sampling strategy. Developing a surrogate model for predicting structures' dynamic responses is challenging because it involves high-dimensional inputs and outputs. For this purpose, a novel surrogate model based on cutting-edge deep learning architectures specialized for capturing temporal relationships within time-series data, namely Long-Short term memory layer and Transformer layer, is designed. After being properly trained, the surrogate model could be utilized in place of the finite element method to evaluate structures' responses without requiring any specialized software. On the other hand, the importance sampling is adopted to reduce the number of calculations required when computing the failure probability by drawing more relevant samples near critical areas. Thanks to the portability of the trained surrogate model, one can integrate the latter with the Importance sampling in a straightforward fashion, forming an efficient framework called TTIS, which represents double advantages: less number of calculations is needed, and the computational time of each calculation is significantly reduced. The proposed approach's applicability and efficiency are demonstrated through three examples with increasing complexity, involving a 1D beam, a 2D frame, and a 3D building structure. The results show that compared to the conventional Monte Carlo simulation, the proposed method can provide highly similar reliability results with a reduction of up to four orders of magnitudes in time complexity.

• Smart Structures and Systems
• /
• 제33권4호
• /
• pp.263-279
• /
• 2024

With the increasing diversity of internet media, available video data have become more convenient and abundant. Related video data-based research has advanced rapidly in recent years owing to advantages such as noncontact, low-cost data acquisition, high spatial resolution, and simultaneity. Additionally, structural nonlinearity extraction has attracted increasing attention as a tool for damage evaluation. This review paper aims to summarize the research experience with the recent developments and applications of video data-based technology for structural nonlinearity extraction and damage evaluation. The most regularly used object detection images and video databases are first summarized, followed by suggestions for obtaining video data on structural nonlinear damage events. Technologies for linear and nonlinear system identification based on video data are then discussed. In addition, common nonlinear damage types in disaster events and prevalent processing algorithms are reviewed in the section on structural damage evaluation using video data uploaded on online platform. Finally, a discussion regarding some potential research directions is proposed to address the weaknesses of the current nonlinear extraction technology based on video data, such as the use of uni-dimensional time-series data as leverage to further achieve nonlinear extraction and the difficulty of real-time detection, including the fields of nonlinear extraction for spatial data, real-time detection, and visualization.

• 응용통계연구
• /
• 제28권5호
• /
• pp.871-883
• /
• 2015

벡터자기회귀모형은 다차원의 시계열 자료간의 선형종속 관계를 연구하는데 효율적인 모형이다. 하지만 차원이 높아질 경우 추정해야할 모수가 급격히 증가하여 추정이 불안정해지고 예측력의 저하 및 해석의 어려움을 동반하는 문제를 가지고 있다. 이를 보완하기 위해서 많은 계수를 0으로 두는 희박벡터자기회귀모형이 제안되었고 고차원 시계열 분석에서 유용함이 밝혀졌다. 이 논문에서는 희박벡터자기회귀모형 추정에 있어서 어떠한 계수를 0으로 두어야 하는지를 판단해주는 한 쌍의 변수에 대한 상관 정도를 추정해주는 커플링 측도를 제안한다. 먼저 이 논문에서는 부분 스펙트럼 일관성에 기반을 둔 커플링 측도를 사용한 변수 선택의 경우 다른 변수의 효과를 제거한 잔차에 기반을 두었기에 좋은 효율성을 보임을 밝힌다. 하지만 부분 스펙트럼 일관성의 경우 벡터자기회귀모형 계수의 비대칭성을 고려하지 못한다는 단점이 있어 이를 보완하고자 필터링을 통해 다른 변수의 효과를 제거한 잔차에 기반을 둔 동시에 비대칭성을 가지는 커플링 측도들, 필터링된 잔차를 이용한 교차 상관성과 그래인저 인과관계를 제안한다. 모의실험을 통해 우리가 제안한 방법론들이 두터운 꼬리를 가지거나 높은 차수의 희박벡터자기회귀모형의 경우에도 매우 정확하게 0이 아닌 변수를 선택함을 보인다.

• 한국환경과학회:학술대회논문집
• /
• 한국환경과학회 2003년도 International Symposium on Clean Environment
• /
• pp.73-78
• /
• 2003

In this research, we implement Realtime Air Diffusion Prediction System which is a parallel Fortran model running on distributed-memory parallel computers. The system is designed for air diffusion simulations with four-dimensional data assimilation. For regional air quality forecasting a series of dynamic downscaling technique is adopted using the NCAR/Penn. State MM5 model which is an atmospheric model. The realtime initial data have been provided daily from the KMA (Korean Meteorological Administration) global spectral model output. It takes huge resources of computation to get 24 hour air quality forecast with this four step dynamic downscaling (27km, 9km, 3km, and lkm). Parallel implementation of the realtime system is imperative to achieve increased throughput since the realtime system have to be performed which correct timing behavior and the sequential code requires a large amount of CPU time for typical simulations. The parallel system uses MPI (Message Passing Interface), a standard library to support high-level routines for message passing. We validate the parallel model by comparing it with the sequential model. For realtime running, we implement a cluster computer which is a distributed-memory parallel computer that links high-performance PCs with high-speed interconnection networks. We use 32 2-CPU nodes and a Myrinet network for the cluster. Since cluster computers more cost effective than conventional distributed parallel computers, we can build a dedicated realtime computer. The system also includes web based Gill (Graphic User Interface) for convenient system management and performance monitoring so that end-users can restart the system easily when the system faults. Performance of the parallel model is analyzed by comparing its execution time with the sequential model, and by calculating communication overhead and load imbalance, which are common problems in parallel processing. Performance analysis is carried out on our cluster which has 32 2-CPU nodes.