Modelling of traffic delays at roundabout using vissim microsimulation model

Abstract

Delay is one of the significant parameters used for the operational performance evaluation of roundabouts. Even though there are various methods for estimating delays at the roundabout, the results can differ from actual delays on the field for Malaysian traffic conditions. Most of the existing models, such as HCM 2010, do not consider the impact of the geometric parameters on delays, besides the driver behaviour is different. Therefore, this study used the VISSIM microsimulation model to assess the impacts of traffic and geometric characteristics on delays at roundabouts. Video recording technique was used to collect data at Senai roundabout such as the entry flow rate, circulating flow rate, critical gap, follow-up headway, vehicle desired speed, and delay. The data collected were used to develop the base model scenario in VISSIM. The calibration and validation of the base model were then conducted using entry traffic flow, circulating traffic flow, and delays measured in the field. In VISSIM, three basic scenarios are defined, namely, the central island diameter (Di= 60m, 70m, and 80 m), entry width (We= 5m, 9m, and 12 m), and the circulating roadway width (Wc= 6 m, 9m, and 12m). For each scenario case, additional scenarios were created utilizing different traffic flow rates (350-2000 veh/hr), representing low, medium, and high traffic flow rates; as a result, 351 scenarios are developed in all. Based on the outcomes of the 351 scenarios, an exponential mathematical model is developed for estimating roundabout delay using the regression analysis technique. The entry flow rate and the circulating flow rate are found to have a strong positive correlation with delay. This implies that an increase in the entry flow rate or the circulating flow rate increases the delay. In terms of geometric parameters influence on the delay, the entry width and circulating roadway width exhibit a strong negative correlation with delay (p-value < 0.05). In other words, increasing the width of the entry or circulating roadway leads to delay reduction. In contrast, the central island diameter has no significant impact on the delay (p-value > 0.05); thus, this parameter is omitted from the final model. With an R2 of 0.79, the derived model is statistically significant at a 95% confidence level, which implies that traffic flow and the geometric parameters account for an approximate of 79% variation in the delay. Moreover, the value of the coefficient of multiple correlations (R = 0.89) affirmed that the relationship between delay and the combined influence of the independent variables is significantly strong. Therefore, this study would serve as a more accurate guide for determining the acceptable degree of service for Malaysian roundabouts to justify related expenditures. It is advised that the proposed model be considered into Malaysia's guidance for performance analysis on at-grade junctions, particularly roundabouts.