Microscopic Simulation Model for Motorcycle Dominated Networks: A Case Study of a VISSIM Simulation Model for a Mixed Traffic Corridor

Abstract

Abstract Microscopic traffic simulation packages, which were invented for car-based environments, have also been used for motorcycle-based environments. Because the interactions between several transport modes (e.g., buses, cars, and motorcycles) are different in these two environments, default parameters relating to these modes (e.g., acceleration, deceleration, speed) should be modified in the simulation of motorcycle-dominated networks. However, there seems to be very little evidence on the modification of these parameters in microscopic traffic simulation for these networks. To fill this gap in the literature, this research develops a VISSIM model (as one of the most appropriate microscopic simulation packages) for motorcycle-dominated mixed transport, which includes data collection, creation of a VISSIM model, and calibration and validation processes. For parameters which cannot be observed on-site, some values for each parameter are chosen based on the default values in VISSIM or local conditions. These values are then run in VISSIM models and tested in the calibration and validation processes to choose the values that best describe vehicle movements. By contrast, parameters that might be collected on-site are statistically analysed and inserted directly in to VISSIM. Acceleration and deceleration of motorcycles, the desired speed of car and bus, the desired speed of motorcycle, minimum lateral distance driving when overtaking vehicles on the same lane, and average standstill distance were applied for developing VISSIM models for a motorcycle-dominated mixed traffic corridor in Hanoi, Vietnam. Acceleration and deceleration surveys were carried out on-site by using the speed gun Stalker ATS. The results of statistical tests prove that the VISSIM models are sufficiently reliable to represent real traffic. Transport modellers can draw on the findings of this study. Desired acceleration and deceleration default values in VISSIM, which are for typical motorcycles used in Europe, should be replaced by local acceleration and deceleration rates. The methodology for developing a VISSIM model can be modified to suit other local motorcycle-dominated mixed traffic conditions.