This study intends to present a traffic node-based and link-based accident prediction models using XGBoost which is very excellent in performance among machine learning models, and to develop those models with sustainability and scalability. Also, we intend to present those models which predict the number of annual traffic accidents based on road types, weather conditions, and traffic information using XGBoost. To this end, data sets were constructed by collecting and preprocessing traffic accident information, road information, weather information, and traffic information. The SHAP method was used to identify the variables affecting the number of traffic accidents. The five main variables of the traffic node-based accident prediction model were snow cover, precipitation, the number of entering lanes and connected links, and slow speed. Otherwise, those of the traffic link-based accident prediction model were snow cover, precipitation, the number of lanes, road length, and slow speed. As the evaluation results of those models, the RMSE values of those models were each 0.2035 and 0.2107. In this study, only data from Sejong City were used to our models, but ours can be applied to all regions where traffic nodes and links are constructed. Therefore, our prediction models can be extended to a wider range.

PURPOSES : For vehicle-alone accidents with a high mortality rate, it is necessary to analyze the factors influencing the severity of roadside fixed-object traffic accidents. METHODS : A total of 313 roadside fixed obstacle traffic accidents, variables related to fixed obstacles, and variables related to road geometry were collected. The estimation model was constructed with data collected using an ordinal probit regression model. RESULTS : Piers, vertical slopes, and distances between roads and objects were the primary causes of increased accident severity. CONCLUSIONS : Countermeasures, such as object removal, relocation, clear zones, frangibles, breakaway poles, etc., are required for accident-prone or dangerous points.

PURPOSES: The purpose of this study is to develop a crash prediction model at signalized intersections, which can capture the randomness and uncertainty of traffic accident forecasting in order to provide more precise results. METHODS: The authors propose a random parameter (RP) approach to overcome the limitation of the Count model that cannot consider the heterogeneity of the assigned locations or road sections. For the model’s development, 55 intersections located in the Daejeon metropolitan area were selected as the scope of the study, and panel data such as the number of crashes, traffic volume, and intersection geometry at each intersection were collected for the analysis. RESULTS: Based on the results of the RP negative binomial crash prediction model developed in this study, it was found that the independent variables such as the log form of average annual traffic volume, presence or absence of left-turn lanes on major roads, presence or absence of right-turn lanes on minor roads, and the number of crosswalks were statistically significant random parameters, and this showed that the variables have a heterogeneous influence on individual intersections. CONCLUSIONS : It was found that the RP model had a better fit to the data than the fixed parameters (FP) model since the RP model reflects the heterogeneity of the individual observations and captures the inconsistent and biased effects.

OBJECTIVES : The objective of this study is to develop a traffic accident model of a roundabout based on the type of land use. METHODS : The traffic accident data from 2010 to 2014 were collected from the“ traffic accident analysis system (TAAS)”data set of the Road Traffic Authority. A multiple linear regression model was utilized in this study to analyze the accidents based on the type of land use. Variables such as geometry and traffic volume were used to develop the accident models based on the type of land use. RESULTS : The main results are as follows. First, the null hypothesis that the type of land use does not affect the number of accidents is rejected. Second, four accident models based on the type of land use have been developed, which are statistically significant (high R2 values). Finally, the total entering and circulating volumes, area of the central island, number of speed breakers, mean number of entry lanes, diameter of the inscribed circle, mean width of the entry lane, area of the roundabout, bus stops, and number of circulatory roadways are analyzed to see how they affect the accident for each type of land use. CONCLUSIONS: The development of the accident models based on the type of land use has revealed that the accident factors at a roundabout are different for each case. Thus, more speed breakers in commercial areas and an inscribed circle of proper diameter in commercial and residential areas are determined to be important for reducing the number of accidents. Additionally, expanding the width of the entry lanes, decreasing the area of the roundabouts in residential areas, and reducing the conflict factors such as bus stops in green spaces are determined to be important.

PURPOSES: The purpose of this study is to verify traffic accident injury severity factors for elderly drivers and the relative relationship of these factors. METHODS: To verify the complicated relationship among traffic accident injury severity factors, this study employed a structural equation model (SEM). To develop the SEM structure, only the severity of human injuries was considered; moreover, the observed variables were selected through confirmatory factor analysis (CFA). The number of fatalities, serious injuries, moderate injuries, and minor injuries were selected for observed variables of severity. For latent variables, the accident situation, environment, and vehicle and driver factors were respectively defined. Seven observed variables were selected among the latent variables. RESULTS: This study showed that the vehicle and driver factor was the most influential factor for accident severity among the latent factors. For the observed variable, the type of vehicle, type of accident, and status of day or night for each latent variable were the most relative observed variables for the accident severity factor. To verify the validity of the SEM, several model fitting methods, including , GFI, AGFI, CFI, and others, were applied, and the model produced meaningful results. CONCLUSIONS: Based on an analysis of results of traffic accident injury severity for elderly drivers, the vehicle and driver factor was the most influential one for injury severity. Therefore, education tailored to elderly drivers is needed to improve driving behavior of elderly driver.

본 연구는 신호교차로 교통사고예측모형 구축 과정 중 일반적으로 제한된 변수의 선정 및 모형의 구축에만 주로 초점이 맞추어진 기존 방법론의 문제점을 개선하고, 자료조사 및 수집 과정에서 발생하는 자료의 불확실한 상태를 인정하면서 자료의 불확실성을 최소화하여 이용할 수 있는 방법론을 개발하는데 연구의 주안점을 두었다. 퍼지추론이론과 신경망이론을 이용한 모형을 구축하였고, 마지막으로 구축된 퍼지추론이론 모형 및 신경망이론 모형과 기존 회귀모형인 포아송 회귀모형간의 통계적인 검증과 실제 Data를 이용한 모형의 적정성을 검토하였다. 모형의 통계적인 검증시 기존모형에 비해 퍼지추론모형과 신경망이론모형이 더 설명력이 높은 것으로 나타났고, 검증에서도 퍼지추론이론과 신경망이론이 적절한 것으로 나타났으며 기존모형보다 사고건수를 예측하는 설명력이 높은 것으로 입증되었다. 본 연구에서 개발된 모형은 계획 및 운영단계에서 신호교차로의 안전성을 측정하는데 활용될 수 있으며, 궁극적으로는 신호교차로에서 교통사고를 줄이는데 기여할 수 있을 것으로 판단된다.

고령화가 진행될수록 고령운전자의 수 역시 증가될 것으로 예상되어 향후 고령운전자에 의한 교통사고는 급증할 것으로 판단된다. 본 연구는 고속도로 교통사고 발생시 고령층과 비고령층의 구분에 따른 교통사고 특성을 분석하였다. 분석결과 비고령층에 비해 고령층에 작용하는 영향요인이 다르게 나타남을 알 수 있었다. 로짓모형을 통해 비고령층과 고령층의 Odds Ratio를 분석하여 사고영향요인에 따른 차이점을 알아보았으며 고령운전자의 사고 분석 모형을 개발하였다. 비고령층에 비해 곡선구간 및 절토구간, 노면의 습기상태일 때 고령운전자의 사고위험이 높은 것으로 나타났다.

일반도로구간에서의 사고발생건수는 2000년부터 2006년까지 감소추세를 보이는 반면 교차로에서의 교통사고건수는 현재까지 꾸준하게 증가하고 있기 때문에 교차로에서의 안전성을 증대시키기 위한 노력이 절실히 필요한 실정이다. 본 연구에서는 신호교차로에서의 도로조건, 교통조건, 교통운영상의 조건 등을 분석하여 교차로의 설계 안전성을 저해시키는 요인들을 찾아내고, 그 요인들과 사고와의 상관관계를 이용하여 지방부 4지 신호교차로의 안전성 판단을 위한 사고예측모형을 개발하고자 한다. 또한 사전에 위험요소를 제거하여 교차로에서의 안전성 평가를 위한 가이드라인을 제시함으로써, 교차로에서의 안전성을 높이는데 그 목적이 있다. 본 연구는 교차로에서의 사고분석을 위하여 비선형 회귀분석을 통해 사고모형을 개발하였고, 이러한 모형들을 이용하여 차대차 사고에 영향을 미치는 주요 설명변수들에 대한 분석을 시도하였다. 모형분석결과, 포아송회귀분석(Poisson Regression)이 모형개발에 가장 적합한 것으로 나타났으며, 사고에 영향을 미치는 변수로는 좌회전 전용차로, 횡단보도, 제한속도, 조명시설, 교차각, ADT 등으로 분석되었다.

본 연구는 도로기하구조 요인과 교통사고간의 관계를 규명하기 위하여 CART분석을 이용하여 전국의 4차로 국도를 대상으로 교통사고예측모형을 개발하고, 다중회귀모형, 확률회귀모형과 CART분석모형을 비교 분석하여 개발한 모형의 적합도를 검증하였다. 연구결과로는 첫째, 변수간의 복합적인 상호관계를 설명할 수 있는 CART분석을 이용하여 국도의 교통사고 예측모형을 개발하고 도로기하구조 요인에 따라 표준교통사고율을 의미하는 교통사고발생도표를 제시하였다. 둘째, CART분석모형에 근거하여 교통사고 발생률에 큰 영향을 미치는 도로기하구조 요인이 구간거리(km), 횡단보도폭(m), 횡단길어깨(m), 교통량 순으로 나타났다. 셋째, CART분석모형의 적합도 검증결과, CART분석모형이 실제교통사고율을 타 모형에 비해 전반적으로 잘 묘사하고 있었으나, 각 모형별로 교통사고율의 크기에 따라 교통사고율이 비교적 낮은 구간에서는 다중회귀모형이, 평균이상의 교통사고율을 나타내는 구간에서는 포아송 회귀모형의 예측력이 높았으며, CART분석모형은 교통사고율의 크기와 상관없이 우수한 예측력을 보였다. 넷째, 도출된 교통사고발생도표는 도로기하구조 조건에 따른 표준교통사고율을 제시해주기 때문에 도로설계 시에 안전한 기하구조 설계요소 선정기준을 제시 할 뿐만 아니라, 교통사고 잦은 지점개선사업추진 시 사업의 우선순위를 판단할 수 있는 기준을 제시하는 등 정책적 활용도가 매우 높을 것으로 판단된다.

Traffic accidents increase with the increase of the vehicles in operation on the street. Especially big traffic accidents composed of over 3 killed or 20 injured accidents with the property damage become one of the serious problems to be solved in most of the cities. The purpose of this study is to build the discrimination model on big traffic accidents using the Quantification II theory for establishing the countermeasures to reduce the big traffic accidents. The results are summarized as follows. 1)The existing traffic accident related model could not explain the phenomena of the current traffic accident appropriately. 2) Based on the big traffic accident types vehicle-vehicle, vehicle-alone, vehicle-pedestrian and vehicle-train accident rates 73%, 20.5% 5.6% and two cases respectively. Based on the law violation types safety driving non-fulfillment center line invasion excess speed and signal disobedience were 48.8%, 38.1% 2.8% and 2.8% respectively. 3) Based on the law violation types major factors in big traffic accidents were road and environment, human, and vehicle in order. Those factors were vehicle, road and environment, and human in order based on types of injured driver’s death. 4) Based on the law violation types total hitting and correlation rates of the model were 53.57% and 0.97853. Based on the types of injured driver’s death total hitting and correlation rates of the model were also 71.4% and 0.59583.