查看文章

This paper presents a behavioral trajectory planning for efficient motion planning in urban environments. For the autonomous driving in these environments, there are various complicated driving situations because of many combinations of road types, multiple lanes, traffic rules, and static and dynamic obstacles. Moreover, since these are changed dynamically, real-time trajectory planning is more intricate. To overcome this problem, we proposed behavioral trajectory planning which generates a macro-scale trajectory to define a spatiotemporal boundary of vehicle motion. This planner focuses on searching for a driving maneuver considering lane structure and traffic flow. As a result, this trajectory can reduce a searching area to generate a motion-level trajectory. To utilize this advantages, we apply a numerical optimization method for a local trajectory. The proposed algorithm was implemented and evaluated by …