http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
머신러닝을 사용한 기상 상태 기반 교통사고 위험도 예측 시스템
신수연(Su-Yeon Shin),손수민(Su-Min Sohn),김동국(Dong-Guk Kim),이기환(Ki-Hwan Lee),오병우(Byoung-Woo Oh) 한국정보기술학회 2021 Proceedings of KIIT Conference Vol.2021 No.6
본 논문은 기상 상태에 따른 교통사고 위험도를 예측하는 시스템을 제안한다. 도로교통공단의 교통사고 분석 시스템에 따르면, 기상 상태에 따른 교통사고 건수의 확연한 차이를 확인할 수 있다. 시스템에서 사용하는 예측 모델의 학습 정보로 교통사고 정보, 기상 정보를 활용한다. 모델에 데이터를 학습하여 실시간 기상 정보와 도로명 정보를 입력했을 때, 교통사고 위험도를 예측하여 반환한다. 또한, 사용자가 실시간으로 교통사고 위험도 예측 기능을 사용할 수 있도록 REST API 형태의 시스템을 제안한다. 사용자는 현재 위치의 도로명 데이터로 교통사고 위험도 데이터를 요청하여 수신한다. 사용자가 교통사고 위험도 정보를 확인하고 미리 사고에 대비함으로써 교통사고 발생률의 감소를 기대할 수 있다. This paper proposes a system that predicts traffic accident risk level according to weather conditions. According to the traffic accident analysis system of the Road Traffic Authority, it is possible to confirm a marked difference in the number of traffic accidents according to weather conditions. Traffic accident information and weather information are used as learning data for predictive models used in the system. After the data is learned into the model and real-time weather data and road name data are input, traffic accident risk levels are predicted and returned. In addition, we propose a system in the form of a REST API that allows users to use the traffic accident risk level prediction function in real time. The user requests the road junction data of the current location and receives the traffic accident risk level data. Users can check traffic accident risk level data and expect a reduction in traffic accident incidence in preparation for accidents in advance.
유역특성에 따른 하천에서의 존재형태별 질소 분포 특성 비교
박지형 ( Ji Hyoung Park ),손수민 ( Su Min Sohn ),김용석 ( Yong Seok Kim ) 한국물환경학회 2014 한국물환경학회지 Vol.30 No.5
The temporal and spatial analyses of total nitrogen (TN) fractionation were conducted in order to understand 1) total nitrogen components in streams and 2) their patterns in rainy and dry seasons. The result showed that the concentration of nitrogen components in stream water was lower in non-urban area and getting higher in urban area. Dissolved total nitrogen (DTN) was 95~97.7% of total nitrogen in streams, and the proportion of dissolved organic nitrogen (DON) and ammonia nitrogen (NH3-N) was higher with increasing urban area. The concentration of total nitrogen and nitrate nitrogen (NO3-N) were highest in winter among four seasons. The result was showed that concentration of NH3-N was same variation as concentrations of TN and NO3-N in urban-rural complex and urban areas, except rural areas. During rainy season, concentrations of particulate organic nitrogen (PON) and NH3-N increased in rural areas and decreased in both urban-rural complex and urban areas. Correlation between total nitrogen components and land uses was positively correlated with site > paddy, and negatively correlated with forest. The variation of total nitrogen concentration was determined by NO3-N in non-urban areas, by NO3-N and NH3-N in urban-rural complex and by NH3-N in the urban areas.