一种基于联邦学习的路径规划方法、设备及介质
By using federated learning for path planning, vehicles train lightweight models in the cloud and perform verification and control on the vehicle, solving the problems of dynamic adaptability and resource consumption in traditional path planning methods, and achieving lightweight model deployment and resource optimization.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHANDONG UNIV OF SCI & TECH
- Filing Date
- 2026-05-12
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technology path planning methods suffer from poor dynamic adaptability, difficulty in lightweight model deployment, and excessive consumption of communication and computing resources.
Using a federated learning approach, the vehicle extracts a structured dataset locally and uploads it to the cloud. The cloud trains a lightweight global path planning model and distributes it to the vehicle. The vehicle performs decoupled verification and dynamic conflict detection in a lightweight digital twin. It generates vehicle control commands by combining real-time sensor data and records decision data to upload to the cloud to update the model.
It enables lightweight deployment of models on resource-constrained vehicles, improves the system's adaptability to dynamic and complex environments, optimizes the use of communication computing resources, and ensures dynamic adaptability and resource balance.
Smart Images
Figure CN122170915B_ABST