Traffic injury prevention · 2025
Nonlinear effects of built environment on road traffic safety: A multi-scale perspective.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Wu Na, Zhang Hengming, Yang Suhe, Pan Xiaofeng
Аннотация
OBJECTIVE: This study aims to address the research gaps regarding the nonlinear effects and two-way interaction effects of multi-scale built environment features on road traffic safety.
METHODS: Using data from online ride-hailing accidents in three Chinese megacities, this research applied a gradient boosting decision trees (GBDT) model to analyze the nonlinear relationships between multi-scale built environment features and road traffic safety. The built environment was characterized by three scales: road attributes, Points of Interest (POIs), and street view images (SVIs). The analysis focused on both the main effects and two-way interaction effects of these features.
RESULTS: The results indicate that SVI features have the highest combined contribution, accounting for 74.08% totally, followed by a combined contribution of 13.77% from POIs. Road characteristics had the least combined contribution to traffic accidents, accounting for 12.16%. Only the parking lot can decrease (but still slightly) the traffic accident risk of bus station, which means that the existence of parking lots near a bus station can decrease the probability of occurrence of traffic accidents. While other two-way interactions would increase traffic accident risks.
CONCLUSIONS: First, drivers' visual information (captured via SVIs) emerged as the most critical factor, contributing 74.08% to road safety outcomes. Second, tunnels, primary roads, residential roads, all POI categories, guardrails, and traffic signals were identified as significant hazards. Moreover, SVI features and road class exhibited pronounced nonlinear effects on safety. Additionally, significant interaction effects were observed between these variables, indicating that their combined influence on safety is more complex than individual effects alone.
METHODS: Using data from online ride-hailing accidents in three Chinese megacities, this research applied a gradient boosting decision trees (GBDT) model to analyze the nonlinear relationships between multi-scale built environment features and road traffic safety. The built environment was characterized by three scales: road attributes, Points of Interest (POIs), and street view images (SVIs). The analysis focused on both the main effects and two-way interaction effects of these features.
RESULTS: The results indicate that SVI features have the highest combined contribution, accounting for 74.08% totally, followed by a combined contribution of 13.77% from POIs. Road characteristics had the least combined contribution to traffic accidents, accounting for 12.16%. Only the parking lot can decrease (but still slightly) the traffic accident risk of bus station, which means that the existence of parking lots near a bus station can decrease the probability of occurrence of traffic accidents. While other two-way interactions would increase traffic accident risks.
CONCLUSIONS: First, drivers' visual information (captured via SVIs) emerged as the most critical factor, contributing 74.08% to road safety outcomes. Second, tunnels, primary roads, residential roads, all POI categories, guardrails, and traffic signals were identified as significant hazards. Moreover, SVI features and road class exhibited pronounced nonlinear effects on safety. Additionally, significant interaction effects were observed between these variables, indicating that their combined influence on safety is more complex than individual effects alone.