Modeling driving behavior and emulating driver control strategy is one of essential parts for risk assessment and auxiliary control in the driver assistance system and vehicle active safety system. In this paper, we propose a dynamic lane-change model which reflects the driver control strategy of adjusting longitude and lateral acceleration. The proposed model is intuitive and can clearly describe the habit and randomness of the lane-change behavior with limit parameters. Especially, by using the proposed model and method driver can choose the different driving scenes, such as slow or sudden, to generate different lane-change trajectories. Compared with the conventional polynomial lane-change path planning, the dynamic model can generate the feasible lane-change trajectories and show the good performance of approximating the real lane-change trajectories.
Generating lane-change trajectories using the dynamic model of driving behavior
2011
6 Seiten, 31 Quellen
Aufsatz (Konferenz)
Englisch
Learning Lane Change Trajectories from On-Road Driving Data
British Library Conference Proceedings | 2012
|