Abstract

Models that predict the operating speed (OS) in terms of highway geometry assume that drivers are homogenous individuals. However, road users build the way they drive based on their behaviour and skills, thereby configuring a so-called driving style. The state-of-the-art literature defines at least eight driving styles that affect the OS of drivers, but OS-geometry models do not consider this aspect. The purpose of this work was to study the effect of driving styles on OS in crest vertical curves (CVCs) of two-way highways. To this end, 16 drivers were selected and classified according to their driving style, based on the multidimensional driving style inventory. Subsequently, the circulation speed of each driver was tested on 40 CVCs through a 10 Hz Global Positioning System logger installed in each vehicle. New OS-vertical geometry models were developed and classified according to driving style. It was concluded that models that do not consider driving styles conceal the sensitivity of the OS to the vertical geometry of highways. Additionally, among the different driving styles there is an OS difference of up to 8 km/h. Therefore, including the driving style variable in OS models of two-way roadways is a relevant factor.