Calibration of Acceleration-Based and Multi-Anticipative Car-Following Models by NGSIM Trajectory Data

Li Jun Sun,Xi Qun Chen,Wei Jun Xie,Xinmiao Yang
DOI: https://doi.org/10.1061/41127(382)192
2010-01-01
Abstract:Recently, the Next Generation Simulation (NGSIM) has attracted many researchers' interests because of its representative data and rich contents. Based on the NGSIM data, this paper calibrates the multi-anticipative car-following models, including multi-anticipative IDM, multi-anticipative FVDM, MHVD model, and MVD model. Before calibration, this article pre-processes the accelerations and velocities since there are several defects in the original data. The calibration of car-following models and multi-anticipative car-following models is grounded on the Levenberg-Marquardt algorithm; furthermore, the calibration results of car-following models can provide important reference value for researchers, such as the appropriate estimating ranges of model parameters, while the results of calibrating multi-anticipative car-following models not only introduce the parameter estimation ranges but also testify that, in reality, the multi-anticipative models are more suitable and two preceding vehicles can be a better choice.
What problem does this paper attempt to address?