Understanding vehicles’ movement in complex environments becomes cruical with the fast development of connected automated vehicles (CAVs). Current microscopic traffic flow models lack consideration for vehicle dynamics and complex road topologies. This study develops a model addressing these issues, retrieving vehicles’ maneuvers and predicting vehicles’ two-dimensional motion. It introduces a two-dimensional intelligent driving model utilizing steering angle and acceleration as control inputs. Intricate road topologies are represented with potential fields and a virtual boundary to capture the heterogeneous environment’s complexity. A driving potential field model is also developed for off-ramp areas. Model parameters are optimized with the dynamic time warping (DTW) and particle swarm optimization (PSO) algorithm. Furthermore, model predictive control (MPC) enhances the realism of the model’s output. Field validation results demonstrate that the proposed models can accurately describe vehicles’ two-dimensional movement in intricate environments, offering valuable support for research on heterogeneous traffic flows for CAVs and CVs.
Intelligent driving model considering vehicular dynamics and heterogeneous road environments
J. ZHANG AND H. QI
TRANSPORTATION LETTERS
Transportation Letters ; 17 , 1 ; 119-134
02.01.2025
16 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Two-Dimensional Intelligent Driver Model with Vehicular Dynamics
British Library Conference Proceedings | 2022
|VEHICULAR DRIVING ASSISTANCE SYSTEM WITH ENHANCED ROAD CURVE MANAGEMENT
Europäisches Patentamt | 2024
|Two-Dimensional Intelligent Driver Model with Vehicular Dynamics
SAE Technical Papers | 2022
|Europäisches Patentamt | 2024
|