A Study On Context Sensitive Highway Design Based On Improved Operating Speed Prediction Methods in National Roads

Highway design speed is a very important design element which determines highway design level. When determining highway design speed, one would estimate it utilizing the most likelihood of design speed and vehicle operating speed relationship. Existing operating speed prediction models only include highway geometric characteristics and their impacts on speed, which usually can not consider the impact of highway design speed on surrounding roadway environment and land use pattern. If this happens, excessive highway construction cost and huge environmental impact can occur. In this research project, a new vehicle operating speed prediction model was developed which can reflect the effect of surrounding roadway environment into vehicle speed prediction. The followings are the research findings : Firstly, highway terrain types and land use pattern on national roads were classified and integrated into drivers' visual recognition pattern. This was performed using a data management software. Secondly, the developed highway terrain types and land use pattern were related to vehicle speeds and it was found that there were significant statistical differences among vehicle speed for each different terrain and land use pattern. Thirdly. the General Linear Model analysis was employed to analyze the effects of highway geometric features, terrain types, and land use patterns. For two-lane highway and four-lane highway tested in this research project, it was found that R squares were 0.67 and 0.85, respectively. Additionally an optimal highway design speed range table, based on this research project. was proposed for practical use. This table can be reliably used on South Korean national road design, but discretion is required for applying this table to other types of highways including provincial roads and municipal roads.