Adaptive Trajectory Generation of Autonomous Vehicle in Urban Environments

Adaptability of autonomous vehicle to road conditions is important to make driving safely in a mixed environment of autonomous vehicle and human driven vehicles operating on the same road. In this paper, we consider an adaptive trajectory generation methodology based on detected information from sensors to deal with obstacles break into driving path. Rule-based conditions and system architecture are constructed based on vehicle dynamic and detected information. Different scenarios including lane change, roadside stop of human drivers, and cut-in situation in urban environments are validated via simulation study. Result shows that autonomous vehicle adaptively avoids the collision well with obstacles by means of lane-change or braking based on proposed manner.