http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
배광수 ( Bae Kwang-soo ),이승철 ( Lee Seung-cheol ),이호원 ( Lee Hoe-won ) 한국도로교통공단 2018 교통안전연구 Vol.37 No.-
Service users are confused about the different criteria such as “term” and “speed range” for differentiating the traffic condition of traffic information service provider. Traffic congestion depends on a driver’s psychology of expectation, and hence, formulating a clear definition regarding traffic information is difficult. Therefore, in this study, we used various methods to prepare a reasonable basis of the method of expressing traffic information. These methods included user questionnaire surveys, Korea Highway Capacity Manual (KHCM), and data mining. Subsequently, we presented improvements such as traffic condition classification per road, the speed range of a traffic condition, and expression term of a traffic condition. It is expected that based on the findings of this study, users can clearly understand traffic information and access it systematically. Another expectation is that users can use the result and approach of this study to distinguish traffic conditions in a quantitative manner.
이항로지스틱 회귀모형과 MLP 신경망을 이용한 신호위반사고 위험교차로 예측모형 개발 - 도시부 4지 교차로를 중심으로 -
배광수 ( Bae Kwang-soo ),박정순 ( Park Jeong-soon ),안계형 ( Ahn Gye-hyeong ),이승철 ( Lee Seung-chul ) 한국도로교통공단 2017 교통안전연구 Vol.36 No.-
RLR(Red-Light-Running) crash is one of the most frequent traffic accident caused by violation of traffic regulation. Despite various measures, RLR crash fatalities have not been decreased in recent decade. Considering RLR crash is caused by various factor at signalized intersection, such as road environmental factor, traffic signal operational factor and human factor, systematic analysis of the cause of RLR crash is very important in establishing traffic safety countermeasures. Therefore, in this study we used a total of 4,221 RLR crash data that occurred in the last three years at 73 intersections of Gyeonggi-do in order to develop RLR crash prediction model and derive accident induction factors. The model construction result, the binominal logistic regression model and the MLP neural network model were found to have high statistical suitability. In addition, it was found that the prediction classification accuracy of the RLR crash danger intersection also shows excellent performance for the two models(binominal logistic 89.0% ↔ MLP neural network 87.7%). Utilizing the development model, we analyzed the main induction factors of RLR crash. As a result, the main factors that had a major influence on the severity of RLR crash were led by traffic volume, yellow signal time, intersection area and so on.
단계시공에 의한 보강콘크리트벽체를 갖는 보강토 옹벽 적용사례
원명수(Myoung-Soo Won),배광수(Kwang-Soo Bae),최용훈(Yong-Hun Choi),박지승(Ji-Seung Park) 한국지반신소재학회 2012 한국토목섬유학회 학술발표회 Vol.2012 No.11
Geosynthetic Reinforced Soil Retaining Wall (GRS RW) with Full-Height Rigid(FHR) Reinforced Concrete Facing by Staged Construction Procedures is reported with the fact that the stability is outstanding, expecially in earthquake. Recently, as the railway construction is getting increase, the interest of GRS RW-FHR is becoming larger day by day. Under these background, the present paper introduces a case which GRS RW-FHR developed with domestic technique is applied in field.
기용걸(Ki,Yong-Kul),배광수(Bae,Kwang-Soo),안용주(Ahn,Yong-Ju),안계형(Ahn,Gye-Hyeong) 한국IT서비스학회 2014 한국IT서비스학회 학술대회 논문집 Vol.2014 No.추계
교통사고로 인한 인명피해를 줄이기 위해서는 운전자에게 교통안전정보(교통사고 다발지점 등) 및 돌발정보(사고, 공사 등)가 효율적으로 제공되어야 한다. 그러나 기존의 교통안전정보 제공체계는 정보수집 및 제공기능이 부족하여 운전자에게 교통안전정보를 적절히 제공하지 못하고 있는 실정이다. 이를 개선하기 위해 중앙교통정보센터에 교통안전정보 관리시스템을 구축하고, 전국의 교통경찰 및 교통안전시설물 관리자들이 보다 효과적으로 교통안전정보 및 돌발정보를 수집, 관리, 제공할 수 있는 교통안전정보 제공서비스 개선방안을 제안하였다. 2013년 말에 본 논문에서 제안한 방식으로 중앙교통정보센터에 교통안전정보 관리시스템을 개발 및 구축하였으며, 전국 250여개 경찰서의 교통경찰들을 중심으로 교통안전 및 돌발상황 관련 최신정보가 수집 및 제공되고 있다.