• Journal of Information Processing Systems
• /
• v.15 no.5
• /
• pp.1156-1170
• /
• 2019

In recent years, the problem of data drifted of the smart grid due to manual operation has been widely studied by researchers in the related domain areas. It has become an important research topic to effectively and reliably find the reasonable data needed in the Supervisory Control and Data Acquisition (SCADA) system has become an important research topic. This paper analyzes the data composition of the smart grid, and explains the power model in two smart grid applications, followed by an analysis on the application of each parameter in density-based spatial clustering of applications with noise (DBSCAN) algorithm. Then a comparison is carried out for the processing effects of the boxplot method, probability weight analysis method and DBSCAN clustering algorithm on the big data driven power grid. According to the comparison results, the performance of the DBSCAN algorithm outperforming other methods in processing effect. The experimental verification shows that the DBSCAN clustering algorithm can effectively screen the power grid data, thereby significantly improving the accuracy and reliability of the calculation result of the main grid's theoretical line loss.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.12
• /
• pp.5978-5999
• /
• 2018

Pedestrian detection is a challenging area in the intelligent vehicles domain. During the last years, many works have been proposed to efficiently detect motion in images. However, the problem becomes more complex when it comes to detecting moving areas while the vehicle is also moving. This paper presents a variational optical flow-based method for motion estimation in vehicular traffic scenarios. We introduce a framework for detecting motion areas with small and large displacements by computing optical flow using a multilevel architecture. The flow field is estimated at the shortest level and then successively computed until the largest level. We include a filtering parameter and a warping process using bicubic interpolation to combine the intermediate flow fields computed at each level during optimization to gain better performance. Furthermore, we find that by including a penalization function, our system is able to effectively reduce the presence of outliers and deal with all expected circumstances in real scenes. Experimental results are performed on various image sequences from Daimler Pedestrian Dataset that includes urban traffic scenarios. Our evaluation demonstrates that despite the complexity of the evaluated scenes, the motion areas with both moving and static camera can be effectively identified.

• Smart Structures and Systems
• /
• v.30 no.1
• /
• pp.63-74
• /
• 2022

The deformation coordination between a rock/soil mass and an optical sensing cable is an important issue for accurate deformation monitoring. A stress-controlled triaxial apparatus was retrofitted by introducing an optical fiber into the soil specimen. High spatial resolution optical frequency domain reflectometry (OFDR) was used for monitoring the strain distribution along the axial direction of the specimen. The results were compared with those measured by a displacement meter. The strain measured by the optical sensing cable has a good linear relationship with the strain calculated by the displacement meter for different confining pressures, which indicates that distributed optical fiber sensing technology is feasible for soil deformation monitoring. The performance of deformation coordination between the sensing cable and the soil during unloading is higher than that during loading based on the strain transfer coefficients. Three hypothetical strain distributions of the triaxial specimen are proposed, based on which theoretical models of the strain transfer coefficients are established. It appears that the parabolic distribution of specimen strain should be more reasonable by comparison. Nevertheless, the strain transfer coefficients obtained by the theoretical models are higher than the measured coefficients. On this basis, a strain transfer model considering slippage at the interface of the sensing cable and the soil is discussed.

• Current Optics and Photonics
• /
• v.5 no.4
• /
• pp.370-374
• /
• 2021

Corals are the remains of animals that grow on warm beaches. They have been used as decorative jewels because of their variety of colors, and as medicinal materials for treating cancers, AIDS, and other therapeutic uses because of their chemical elements. Corals are mainly composed of calcium carbonate (CaCO₃) and have many air pores, tens to hundreds of micrometers in size. The refractive indices and absorption coefficients of dried sliced staghorn corals are investigated using terahertz time-domain spectroscopy. The measured values are similar to those for CaCO₃, as expected. It is observed that a sample with a microstructure formed by air pores can guide terahertz waves. The dispersion, effective index, and loss of the guiding modes of coral core surrounded by five triangular air pores are numerically calculated. The simulated spatial distribution of the electric field of the guide mode at 1.25 THz shows the mode to be tightly confined to the core.

• Journal of Korea Multimedia Society
• /
• v.24 no.11
• /
• pp.1552-1559
• /
• 2021

Convolution neural networks (CNNs) show notable performance in image processing and are used as representative core models. CNNs extract and learn features from large amounts of train dataset. In general, it has a structure in which a convolution layer and a fully connected layer are stacked. The core of CNN is the convolution layer. The size of the kernel used for feature extraction and the number that affect the depth of the feature map determine the amount of weight parameters of the CNN that can be learned. These parameters are the main causes of increasing the computational complexity and memory usage of the entire neural network. The most computationally expensive components in CNNs are fully connected and spatial convolution computations. In this paper, we propose a Fourier Convolution Neural Network that performs the operation of the convolution layer in the Fourier domain. We work on modifying and improving the amount of computation by applying the fast fourier transform method. Using the MNIST dataset, the performance was similar to that of the general CNN in terms of accuracy. In terms of operation speed, 7.2% faster operation speed was achieved. An average of 19% faster speed was achieved in experiments using 1024x1024 images and various sizes of kernels.

• The Bulletin of The Korean Astronomical Society
• /
• v.46 no.1
• /
• pp.47.4-48
• /
• 2021

Gravitationally lensed Type Ia supernovae may be the next frontier in cosmic probes, able to deliver independent constraints on dark energy, spatial curvature, and the Hubble constant. Measurements of time delays between the multiple images become more incisive due to the standardized candle nature of the source, monitoring for months rather than years, and partial immunity to microlensing. While currently extremely rare, hundreds of such systems should be detected by upcoming time-domain surveys. Others will have the images spatially unresolved, with the observed lightcurve a superposition of time delayed image fluxes. We investigate whether unresolved images can be recognized as lensed sources given only lightcurve information and whether time delays can be extracted robustly. We develop a method that we show can identify these systems for the case of lensed Type Ia supernovae with two images and time delays exceeding ten days. When tested on such an ensemble the method achieves a false positive rate of ≲5%, and measures the time delays with the completeness of ≳93% and with a bias of ≲0.5% for time delay ≳10 days. Since the method does not assume a template of any particular type of SN, the method has the potential to work on other types of lensed SNe systems and possibly on other transients.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.117-117
• /
• 2021

본 연구에서는 Sentinel-1A/B C-band SAR(Synthetic Aperture Radar) 위성영상을 기반으로 인공신경망(Artificial Neural Network, ANN) 모형을 활용해 금강 유역 상류 40×50 km² 면적에 대한 토양수분을 산정하였다. 10 m 공간 해상도의 Sentinel-1A/B SAR 영상은 8일 간격으로 2015년부터 2019년까지 5년 동안 구축하였고, SNAP(SentiNel Application Platform)을 통해 기하 보정, 방사 보정 및 잡음(Noise) 보정을 수행하고 VV 및 VH 편파 후방산란계수로 변환하였다. ANN 모형 검증자료로 TDR(Time Domain Reflectometry)로 측정된 9개 지점의 실측 토양수분 자료를 구축하였으며, 수문학적 개념인 선행강우를 고려하기 위해 동지점에 대한 강수량 자료를 구축하였다. ANN은 각 지점에 해당하는 토양 속성별로 모델링하고, 전체 기간 및 계절별로 나누어 모의하였으며, 전체 자료의 60%와 40%를 각각 훈련 및 테스트 데이터로 사용하였다. 산정된 토양수분은 상관계수(Correlation Coefficient, R)와 평균제곱근오차(Root Mean Square Error, RMSE)를 활용하여 검증을 수행할 예정이다.

• Steel and Composite Structures
• /
• v.48 no.2
• /
• pp.131-143
• /
• 2023

The CFRP bar was used to achieve more ductile and durable headed-stud shear connectors in composite components. Three series of push-out tests were firstly conducted, including specimens reinforced with pure steel fibers, steel and CFRP bars. The distributed stress was measured by the commercial PPP-BOTDA (Pre-Pump-Pulse Brillouin optical time domain analysis) optical fiber sensor with high spatial resolution. A series of numerical analyses using non-linear FE models were also made to study the shear force transfer mechanism and crack response based on the test results. Test results show that the CFRP bar increases the shear strength and stiffness of the large diameter headed-stud shear connection, and it has equivalent reinforcing effects on the stud shear capacity as the commonly used steel bar. The embedded CFRP bar can also largely improve the shear force transfer mechanism and decrease the tensile stress in the transverse direction. The parametric study shows that low content steel fibers could delay the crack initiation of slab around the large diameter stud, and the CFRP bar with normal elastic modulus and the standard reinforcement ratio has good resistance to splitting crack growth in headed stud shear connectors.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.06a
• /
• pp.818-820
• /
• 2022

포인트 클라우드 데이터는 자율 주행 기술, 가상 현실 및 증강 현실에서 사용될 3차원 미디어 중 하나로 각광 받고 있다. 국제 표준화 기구인 MPEG(Moving Picture Expert Group)에서는 포인트 클라우드 데이터의 효율적인 압축을 위해 G-PCC(Geometry-based Point Cloud Compression) 및 V-PCC(Video-based Point Cloud Compression)의 표준화를 진행 중에 있다. 그 중, G-PCC는 본래 단일 프레임의 압축을 수행하는 정지 영상 압축 방식이지만, LiDAR(Light Detection And Ranging) 센서를 통해 획득된 동적 포인트 클라우드 프레임에 대한 압축의 필요성이 대두됨에 따라 G-PCC 그룹에서는 Inter-EM(Exploratory Model)을 신설하여 LiDAR 포인트 클라우드 프레임의 압축에 관한 연구를 시작하였다. Inter-EM의 압축 비트스트림은 G-PCC 비트스트림과 마찬가지로 효과적인 전송 및 소비를 위해 미디어 저장 포맷인 ISOBMFF(ISO-based Media File Format)으로 캡슐화될 수 있다. 이때, 포인트 클라우드 프레임들은 자율 주행 등의 서비스에 사용하기 위해 시간 도메인뿐만 아니라 공간 도메인을 기반으로도 소비될 수 있어야 하지만, 공간 도메인을 기반으로 콘텐츠를 임의 접근하여 소비하는 방식은 기존 2D 영상의 시간 도메인 기반 소비방식과 차이로 인해 기존에 논의된 G-PCC 캡슐화 방안만으로는 지원이 제한된다. 이에, 본 논문에서는 G-PCC 콘텐츠를 공간 도메인에 따라 소비하기 위한 ISOBMFF 캡슐화 방안에 대한 파일 포맷을 제안하고자 한다.

• Steel and Composite Structures
• /
• v.50 no.4
• /
• pp.429-441
• /
• 2024

This paper studies the free vibration behavior of trapezoidal shaped coupled double-layered graphene sheets (DLGS) system using first-order shear deformation theory (FSDT) and incorporating nonlocal elasticity theory. Two nanoplates are assumed to be bonded by an interlayer van der walls force and surrounded by an external kelvin-voight viscoelastic medium. The governing equations together with related boundary condition are discretized using a mapping-differential quadrature method (DQM) in the spatial domain. Then the natural frequency of the system is obtained by solving the eigen value matrix equation. The validity of the current study is evaluated by comparing its numerical results with those available in the literature and then a parametric study is thoroughly performed, concentrating on the series effects of angles and aspect ratio of GS, viscoelastic medium, and nonlocal parameter. The model is used to study the vibration of DLGS for two typical deformation modes, the in-phase and out-of-phase vibrations, which are investigated. Numerical results indicate that due to Increasing the damping parameter of the viscoelastic medium has reduced the frequency of both modes and this medium has been able to overdamped the oscillations and by increasing stiffness parameters both in-phase and out-of-phase vibration frequencies increased.