For the investigation of driving functions and corresponding components (e.g. sensor and camera systems), automated vehicle manufacturers and related industrial partners rely on simulations. In particular, to investigate advanced vehicle systems in urban areas, simulation setups have to be applied based on realistic traffic scenarios. The presented work outlines a methodology for adapting the longitudinal driving behavior of ambient traffic in microscopic traffic flow simulation based on real driving data. For this purpose, the optimization is performed on the simulation’s car-following model to fit the existing dataset. Accordingly, the desired driving behavior is adapted by adjusting the cumulative acceleration probability. The calibration result obtained a total error of 4.60% between simulation and real driving. After the completed calibration, the results show that the presented methodology is suitable to replicate similar acceleration probabilities for the given dataset.
Calibration of Traffic Flow Simulation Towards Realistic Human Behavior
14.04.2022
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Calibration of Traffic Flow Simulation Towards Realistic Human Behavior
DataCite | 2022
|A Hierarchical Pedestrian Behavior Model to Generate Realistic Human Behavior in Traffic Simulation
ArXiv | 2022
|Towards a realistic optimization of urban traffic flows
IEEE | 2012
|