Application of Fuzzy Quantitative Theory for Crash Rate Prediction at Roadway Sections

摘要

A traffic crash prediction model was developed to accurately estimate roadway section safety. Crash rate of a roadway section was selected as the predicting measure. Moreover, the following seven factors were selected as major influencing factors: number of years of driving, number of lanes, radius of horizontal curves, longitudinal grade, road surface status, type of intersection, and width of road surface. The traffic crash prediction model was established based on fuzzy quantitative theory. To validate the model, crash data were collected at an actual site between 490.900km to 550.789km on a national highway. The result from the model was in agreement with the actual field data. The conclusion is that the model can be applied to other roadway locations.