Navigation method and apparatus
By integrating road network and historical route information through self-supervised training of graph neural networks, the problem of insufficient generalization ability of navigation route planning in existing technologies is solved, and efficient navigation in multiple scenarios is achieved.
CN122408804APending Publication Date: 2026-07-17HUAWEI TECH CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HUAWEI TECH CO LTD
- Filing Date
- 2025-01-15
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
Existing technologies rely heavily on spatiotemporal distribution when planning routes based on historical trajectories, making them unsuitable for spatiotemporal scenarios outside of historical trajectories and resulting in poor generalization ability.
Method used
By referencing massive historical route data to obtain road network features, and using graph neural networks for self-supervised training, road network and historical route information are integrated to improve the generalization ability of road segment features.
Benefits of technology
It improves the applicability and generalization of navigation routes in various scenarios, and enhances the flexibility and relevance of navigation routes in different navigation scenarios.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN122408804A_ABST
Abstract
本申请实施例提供了一种导航方法及装置，涉及导航领域，该方法包括：参考海量的历史路线数据，获得路网数据的路网特征之后，将路网特征输入到图神经网络，以得到路网中多个路段的原始特征；基于路段的原始特征，通过自监督的方式对第一网络训练；基于训练后的第一网络，确定导航路线。可以在原始特征中集成历史路线信息和路网信息。提高了获取的路段特征的泛化能力，可通用各个场景的路线规划。
Need to check novelty before this filing date? Find Prior Art