Microscopic Safety Evaluation and Prediction for Freeway-to-Freeway Interchange Ramps

Abstract

Freeway-to-freeway interchange ramps are critical components of freeway networks. The safety of interchange ramps is a concern because of their complex horizontal and vertical alignment. For a better understanding of the crash mechanisms of interchange ramps, this work built multilevel Poisson–lognormal models to estimate 3-h-interval crash frequencies and built multilevel logistic regression models to predict real-time crash risks. All models were applied to both single-vehicle and multivehicle crashes. In addition, this study explored the feasibility of the use of crash reports to identify roadway surface conditions at study sites. The crash frequency models revealed that the logarithm of 3-h traffic volume and average turning angle were positive significant parameters in estimating single-vehicle crash frequency and that high traffic volume, sag, or downgrade vertical curve increased multivehicle crash frequency. The crash risk models revealed that the average turning angle had a positive impact on single-vehicle crash risk. Multivehicle crash risk increased if lane occupancy increased or interchange ramp vertical alignment was a downgrade. Furthermore, the crash risk estimation models indicated that roadway surface condition was one of the most important parameters: wet roadway surfaces increased the single-vehicle crash ratio by 8.87 and the multivehicle crash ratio by 2.82. This study proves that implementing crash reports is an effective method for providing a study event’s weather information. After the weather information from crash reports is added, 36.8% more studied events get roadway surface condition information, and the accuracy increases by 7.4%.