عنوان مقاله [English]
Four-lane divided highways constitute a significant proportion of Iranian interurban road network. Capacity is one of the key features when examining traffic conditions on this type of roads which is affected by the driving behavior and local conditions governing these roads. An experimental research approach based on site observations was adopted. The objective was to examine and compare the performance of a number of internationally known methods that so far have been used for the capacity analysis of these roads in Iran, and to develop new models, if required. The other objective was to identify the major parameters that would affect the capacity of these roads. For this purpose, 32 suitable sites were identified and data describing their traffic, geometric and surrounding road environment were collected and analyzed. The results indicated that neither of procedures presented in the German, Indonesian and American Highway Capacity Manuals provided a close estimation of the capacity of this road type in Iran. The results provided from the two newly models developed in this study namely, a neural network model and a multivariate linear regression model indicated that these models produced more accurate results in comparison with those procedures, respectively. Finally, the results indicated that roadway width, lateral friction due to the activity and accessibility of the surrounding area, and combined effect of the grade length and steepness are the major factors affecting the capacity of this road types. The developed models could be used for the capacity analysis of this road type in Iran.