PURPOSES : Under the Traffic Safety Act, the installation and management of transportation facilities (facilities and attachments necessary for the operation of transportation, such as roads, railways, and terminals) must take necessary measures to ensure traffic safety, such as enhancing safety facilities. Recently, railway operators have graded the congestion level inside railway stations and vehicles, addressing safety and convenience issues arising from congestion and providing this information to users. However, for bus-related transportation facilities (such as bus stops, terminals, and transfer facilities), criteria and related research for assessing traffic congestion are lacking. Therefore, this study developed a model for the congestion risk factors of four bus-related transportation facilities and proposed criteria for classifying congestion risk levels. METHODS : This study involved selecting congestion risk influence variables for each traffic facility through field surveys, calculating congestion risk index values through evacuation and pedestrian simulations, and constructing a congestion risk influence model based on the ridge model. RESULTS : The factors influencing congestion were selected to include the number of people waiting, effective sidewalk width, and number of bus stops. As a result of developing congestion risk grades, the central bus stops were determined to be in a severe stage if the Average Waiting Time (AWT) was 2.7 or above. Roadside bus stops were considered severe at 4.2, underground metropolitan transit centers at 3.7, and bus terminals at 5.9 or above. CONCLUSIONS : This study can help establish a foundation for a safety management system for congested areas in transportation facilities. When the congestion risk prediction results correspond to cautionary or severe levels, measures that can reduce congestion risk must be applied to ensure the safety of road users.

빈도별 홍수량의 지역분석을 위하여 유역의 지형특성을 독립변수로 이용하는 회귀모형을 검정하였다. 그런데 이들 독립변수들간의 상관관계가 존재할 경우 능형회귀모형이 이용되기도 하는 이 방법은 다중공선성 문제를 극복하는데 적합한 방법으로 알려져 있다. 능형회귀모형을 최적화하기 위해서는 조정변수가 포함되는 비용함수를 최소화하여야 한다. 본 연구에서는 이 최적화를 위하여 유전알고리즘을 이용하였다. 유전알고리즘은 자연 생물의 유전 및 진화과정을 모방한 추계학적 탐색