ANALYSIS OF ROUNDABOUT STOP-LINE DELAYS: EFFECTS OF KINEMATICAL AND BEHAVIOURAL PARAMETERS IN THE SIMULATION PROCESS OF OBSERVED TRAFFIC CONDITIONS

Nowadays, the functional design of a road junction is carried out by the use of the microsimulation models software, but it is important to underline that, despite the great diffusion of these instruments, there has not been the same improvement of appropriate methods in order to validate and calibrate these models. The main target to pursue is to know which and how many input parameters are most significant for the variation of output results that microscopic simulation software is able to provide. In this study, Authors present and complete a calibration procedure between observed performances of a roundabout placed in a city in province of Cosenza (ITALY) and performances obtained by the use of micro-simulation software (VISSIM). Two sets of scenarios different among them only for the traffic flow distribution were analyzed: Free Flow Condition (FFC), from which to derive the average speed profiles along a through movement; and Saturation Flow Condition (SFC), to determine the average stop-line delay along a particular branch. A total of 216+216 scenarios for single-lane roundabouts has been composed and analyzed in order to evaluate the best combinations of kinematic and behavioural (mainly in terms of acceptable gap) parameters in the simulation of roundabouts and to estimate of average percent errors between observed and simulated performances. This study has allowed to delineate some important conclusions about the choice calibration procedures of micro-simulation software used in experimental plan. In particular: a range of approach speed, a range of length of a reduced speed zone for each entry, a definite value of critical gap and the position of desired exiting speed section represent the more revealing parameters that have allowed the best calibration.