드론 영상데이터를 활용한 신호 교차로 세부 상충 위험 구간 식별

초록

Signalized intersection is a complex road section where various modes of transportation, such as pedestrians, bicycles, and vehicles, converge, requiring heightened attention to traffic safety. Traffic crashes occurring at intersections, particularly those involving vulnerable road users, can result in severe injuries. Therefore, identifying risky areas in advance and developing detailed safety management strategies for these locations are crucial to preventing crashes and mitigating their severity. This study proposes a traffic conflict indicator-based micro-hotspot identification methodology to detect detailed high crash risk sections at signalized intersections and design safety interventions considering their infrastructural and traffic environment characteristics. The study derived traffic conflict indicators, including Time to Collision (TTC), Modified TTC (MTTC), and Post Encroachment Time (PET). To determine the most effective conflict indicator and its threshold for representing crash risks, a generalized linear model was constructed, using crash frequency as the dependent variable and traffic conflict indicators as independent variables. By employing this model, micro-hotspots were identified based on the conflict indicator that most effectively captured crash-prone areas. The analysis revealed that MTTC with thresholds of 1.8 seconds and 1.2 seconds, and PET with a threshold of 1.7 seconds, effectively represented high-risk situations at three intersections, respectively. These indicators highlighted micro-hotspots such as intersection approaches and sections where physically unsegregated bicycle lanes were present, exhibiting elevated crash risks. The findings of this study are expected to contribute to the identification of detailed high-risk segments at signalized intersections. Furthermore, they provide a foundation for developing more refined traffic safety management strategies by analyzing roadway geometry, traffic characteristics, and other factors associated with micro-hotspots.

키워드