http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
머신러닝 기법을 활용한 에너지 데이터 분석에 관한 연구
김동주,권성철,문종희,심기도,배문성,Kim, Dongjoo,Kwon, Seongchul,Moon, Jonghui,Sim, Gido,Bae, Moonsung 한국전력공사 2021 KEPCO Journal on electric power and energy Vol.7 No.2
After the spread of the data collection devices such as smart meters, energy data is increasingly collected in a variety of ways, and its importance continues to grow. However, due to technical or practical limitations, errors such as missing or outliers in the data occur during data collection process. Especially in the case of customer-related data, billing problems may occur, so energy companies are conducting various research to process such data. In addition, efforts are being made to create added value from data, which makes it difficult to provide such services unless reliability of data is guaranteed. In order to solve these challenges, this research analyzes prior research related to bad data processing specifically in the energy field, and propose new missing value processing methods to improve the reliability and field utilization of energy data.
의류 수요 정보 예측을 위한 멀티모달 기반 딥 뉴럴 네트워크
김동주(Dongjoo Kim),이민식(Minsik Lee) 대한전자공학회 2022 대한전자공학회 학술대회 Vol.2022 No.11
This paper proposes a deep neural network to estimate the quantitative demand information of clothes displayed in an online shopping mall, using images and other miscellaneous information. In this paper, we construct a clothing demand evaluation database with clothing images collected from an online shopping mall, along with product names, prices, genders, preferred genders, preferred age groups, product numbers, cumulative sales, and product categories. Various modalities are combined in the proposed deep neural network to predict the demand information. To process the images and the text information (product names), we employ ResNet18 and the Multilingual BERT, respectively, in the proposed network. The experimental results show that the proposed multi-modal network shows reliable performance in predicting the demand of clothing.
김동주(Dongjoo Kim) 한국유체기계학회 2006 유체기계 연구개발 발표회 논문집 Vol.- No.-
Numerical simulation of laminar flow over a sphere is conducted to investigate the effect of the Reynolds number on the characteristics of vortex shedding. The Reynolds numbers considered are between 300 and 700, covering unsteady planar-symmetric and asymmetric flows. Results show that the unsteady planar-symmetric flow can be categorized into two different regimes: single-frequency regime and multiple-frequency regime. The former has a single frequency component due to periodic shedding of the same vortex in every shedding cycle, while the latter has multiple-frequency components due to cycle-to-cycle variation in the strength of shed vortices. The multiple-frequency planar-symmetric flow, which is newly found in the present study, occurs at Re=340, 350, and 360 between the single-frequency planar-symmetric and asymmetric flows. On the other hand, the asymmetric flow occurs at Re≥375, where the vortex shed from the sphere shows variation both in strength and azimuthal angle unlike the planar-symmetric flow. Also, it is shown that the breaking of planar-symmetry is closely related to the imbalance between a pair of streamwise vortex in vortical strength.
새만금호 농업용지 방수제 주요지점의 수위 및 유속 수치모의
김동주 ( Dongjoo Kim ),박영진 ( Youngjin Park ),김지성 ( Jisung Kim ),김하집 ( Hajip Kim ) 한국농공학회 2011 한국농공학회 학술대회초록집 Vol.2011 No.-
본 연구는 정부의 새만금 내부개발 기본구상의 변경취지에 부합되고 선도사업으로 추진되는 농업용지구간(54.2km)에 대해 시나리오별 수치모델을 구축하였다. 방수제 시공 현장 여건을 반영한 수치모델을 재현하여 각 공구별로 방수제시공시 세굴이 예측되는 곳에 대한 수치해석을 수행하여 공구별 방수제에 대한 수리학적 검토를 통해 장래 발생 가능한 변화를 예측하고자 하였다. 세굴이 예측되는 곳에 대한 수치해석 결과 100년 빈도 홍수위 조건에서 0.02㎧∼1.27㎧로 모의되어 세굴에 큰 영향을 주는 인자인 유속에 대해 안정적인 것으로 판단된다. 각각 방수제 공구별 시ㆍ종점 부분의 수위와 유속을 검토한 결과 100년 빈도의 홍수량이 유입시 유속분포가 0.02∼1.68㎧로 큰 유속이 발생되지 않는 것으로 분석되어 방수제 시공시 특별한 문제점은 발생되지 않는 것으로 판단되며, 시공 방법과 순서에 따른 방수제 안정성 검토가 앞으로 이루어져야 할 과제이다. 본 연구결과의 결과는 내부 개발 공사에 필요한 수리적인 기초자료를 제공함은 물론 새만금 종합간척사업과 관련하여 수자원 및 환경 등의 각종 정책 결정시 기초 자료로 활용될 것으로 판단된다. This study, the internal development of the Saemangeum basic concept of the changes being promoted as a lead construction and agricultural land works(54.2km) has established a numerical model for the scenario. Inner dikes to the construction site to reflect the following conditions to reproduce the numerical model by each areas during construction inner dikes where scour expected to perform a numerical analysis for the hydraulic review by areas with possible future changes were to predict. Numerical simulation results for scour expected frequency of 100 years in flood conditions is simulated with 0.02㎧∼1.27 ㎧ scour velocity for high-impact factor is considered to be stable. Each start point and end point work area of water proof agent reviewed and flow rate of 100 years flood, the velocity distribution in the influx of a large flow rate of 0.02 ∼1.68 ㎧ occur during construction inner dikes agent are not being evaluated as a special issue does not occur will be considered according to the method and order of construction inner dikes agent stability review suggests that the future need to be made.