Safety is one of the minimum criteria that should be met on the toll road operation. Ideally, the level of safety of highway sections can be predicted even if the highway is still on the design stage. This paper aims to develop mathematical models that could be used to predict the number of accidents, by its level of severity, on inter urban toll road sections based on its average daily traffic per lane and section length. Two yearperiod of traffic and accident data were obtained from Jagorawi, Jakarta-Cikampek, Padaleunyi, and Palikanci toll road operators. Models were developed using the negative binomial regression method. The reluts show that the negative binomial regression gives desirable properties in describing the relationshipbetween the accident frequency and the average daily traffic per lane on each toll road section observed.Keywords: toll road safety, accident model, negative binomial regression.
Copyrights © 2010