This study focused on vulnerable road user crashes induced by mixed traffic at signalized intersections in China. Data on various possible risk factors including road design, signal control, and weather conditions were collected and analyzed accordingly. A count data model was adopted for crash frequency analysis with random matched resampling database to remove the effects of excess zeros on prediction performance. The results indicate that more accurate and reliable parameter estimates and reduced standard errors can be obtained with the random matched resampling method. As results, cycle time of signal, separated right-turn lane (SRL), left-turn waiting area (LWA), and awnings are all significantly related to the risk of vehicle-to-vulnerable road user crashes. This helps to shed light on the development of remedial engineering and traffic management and control measures.
展开▼
