Analysis of Effects of Pavement Marking Retroreflectivity on Traffic Crash Frequency on Highways in North Carolina

摘要: Pavement markings provide useful visual and navigational guidance information to motorists. Current warrants for the application of pavement in United States are based on traffic volume, traveled way width, number travel lanes. To be effective, must visible drivers, particularly at night. The purpose this paper is perform an exploratory analysis determine if a relationship between marking retroreflectivity crash frequency exists. First, models degradation were developed from selected highways North Carolina using artificial neural networks. Monthly estimates levels then appended roadway inventory data. Generalized estimating equations used estimate monthly target frequency. results indicate that regression parameter yellow white edgeline negative, but neither was statistically significant two-lane highway nighttime model. For multilane highways, all significant. as expected. positive.