Motion planning in autonomous driving is a major challenge, as it needs to consider the continuity between the actions output by internal modules and the need for consistency in optimizing driving behavior. Otherwise, it may lead to overly conservative or irrational driving behavior in complex scenarios. To address these issues, we propose a hybrid model predictive motion planner (HMPC) that integrates logical decision-making and motion planning, enabling seamless planning of vehicle motion without semantic decisions or predefined trajectories while extending functionality for external decisions. Testing results reveal that HMPC surpasses the conventional hierarchical MPC motion planner, ensuring better maintenance of the desired speed and greater adaptability to external behavior decision-making modules.
A Seamless Motion Planning Integrating Maneuver Decision Based on Hybrid Model Predictive Control
2023-09-24
2992487 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Maneuver Motion Planning Based on Virtual Target for UAVs
Springer Verlag | 2021
|Maneuver Motion Planning Based on Virtual Target for UAVs
TIBKAT | 2022
|