面向自动驾驶的轻地图化轨迹规划方法
By learning a trajectory behavior dictionary from massive amounts of driving data and employing regional partitioning clustering and hierarchical residual query, the dependence of autonomous driving systems on the centerline of high-precision maps was resolved. This enabled high-precision trajectory planning and multimodal decision-making in mapless environments, improving the generalization ability and functional continuity of autonomous driving systems.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-04-15
- Publication Date
- 2026-07-17
AI Technical Summary
Existing autonomous driving systems rely excessively on the centerline of high-precision maps, resulting in high costs, difficulties in updating and maintenance, drastic drops in functionality and experience, and insufficient generalization capabilities. They also struggle to maintain high accuracy and multimodal decision-making capabilities in mapless environments.
By learning a structured trajectory behavior dictionary from massive amounts of driving data, and using regional partitioning clustering and hierarchical residual query, a trajectory planning method is generated. This method abandons the dependence on the center line of high-precision maps and uses a lightweight expert routing network combined with real-time scene information for trajectory planning.
Maintaining high-precision trajectory planning in unmapped environments enhances the multimodal decision-making capabilities of autonomous driving systems in complex scenarios, eliminates the functional gap between mapped and unmapped areas, and enables autonomous learning and dynamic adjustment of trajectory planning.
Smart Images
Figure CN122408808A_ABST