Modelling Heterogeneous Traffic Performance During Non-Recurrent Congestion (NRC): A Case Study of Tugu Bundaran Sweta Intersection

Abstract

This research delves into the intricate dynamics of traffic performance during non-recurrent congestion (NRC), specifically focusing on the Lebaran Topat festival in Lombok. Utilising sophisticated transport modelling software, namely PTV Vissim, the study systematically models real-world traffic scenarios and calibrates and validates the simulation model’s accuracy. The calibration of driver behaviour parameters, guided by the Wiedemann 74 model, is pivotal in comprehending nuanced vehicle interactions. Data collected from field observation was analysed in preparation for the calibration process, which employed maximum queue length as the parameter and an innovative trial and error method for optimisation. Specific values were assigned for each vehicle type during this calibration, exemplified by an average standstill distance. The validation of the NRC traffic performance model using the chi-square test stands as a testament to its reliability, boasting a chi-square value of 0.12, lower than the critical value of 0.18, at a significance level of 98%. Field assessment shows that during Lebaran Topat, the Level of Service (LOS) for signalised intersections based on the field maximum queue length is F, improved to D after simulated through PTV Vissim.