Traffic Risk Prediction on Ramp Sections based on Multiple Linear Regression
In order to identify potential risks in ramp expressway sections and improve driving safety,a multiple linear regression driving risk prediction model is established in the paper based on field research results,taking the section of a service area in Hunan Province as the research object.Firstly,a simulation model for the research area is established based on the Vissim,and a particle swarm optimization algorithm is used for parameter calibration.Secondly,275 traffic scenarios are designed based on weather conditions,service levels and spatial locations.Secondly,the traffic conflict rate is selected as the risk assessment indicator,and the conflict rates in different scenarios are obtained through traffic simulation.Finally,a risk prediction model for ramp sections based on multiple linear regression is constructed with weather,traffic volume and spatial location as independent variables and conflict rate as dependent variables.The results indicate that under the same weather conditions,the lower the service level,the higher the driving risk on the section.When the visibility is below 100 meters,the risk of road conflict significantly increases.The driving risk is highest within 100 meters upstream and downstream of the junction nose of the expressway and significantly reduced within 300 meters.The driving risks can be mitigated by extending the flexible pile column at the entrance and providing warning and guidance.
NETL
NSTL
万方数据
