Maneuver delegation and planning for automated vehicles at multi-lane road intersections

摘要

Multi-lane intersections set high demands for maneuver planning modules of automated vehicles due to the complex environment as well as the high traffic and maneuver density. Complex interdependencies arise in case the driver is integrated into the automation concept by shared vehicle control on maneuver level. Therefore, we introduce and evaluate different maneuver delegation strategies for multi-lane intersections and describe and quantify their consequences for the vehicle trajectory in a reference intersection scenario. Also, a novel trajectory back propagation approach is presented that allows the determination of fundamental reference points within the intersection that are needed for a comprehensive maneuver- and behavior planning approach that can also be used for fully automated vehicle concepts.