A vehicle-cloud collaborative intelligent driving method, system and medium

CN122924176APending Publication Date: 2026-10-16CHONGQING CHANGAN AUTOMOBILE CO LTD
0 Cites 0 Cited by

Patent Information

Application Number
CN202611157589.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-31
Publication Date
2026-10-16

AI Technical Summary

Technical Problem

[0003]本发明提供了一种车云协同智能驾驶方法、系统及介质，以解决车云协同智能驾驶方法同时存在的响应能力低以及应用成本高的问题

Benefits of technology

[0025]Based on the aforementioned technical means, the cloud continuously receives real-time global features transmitted from vehicles and continuously iterates the historical local features stored in each tile block to achieve crowdsourced updates of road scene data. Simultaneously, differentiated compression storage is performed on the transmitted features, discarding the massive resource consumption caused by long-term storage of original high-dimensional complete features. During compression, the cloud divides features into three categories based on the distance between the vehicle and the tile block for hierarchical storage. The current block completely preserves high-precision original features, ensuring no information loss during high-frequency vehicle access. Adjacent blocks use learnable vector quantization compression, significantly reducing data volume by relying on codebook vectors while retaining sufficient scene details. Far-reaching blocks use coarser-grained compression to maximize the saving of cloud storage space. The hierarchical compression strategy allocates storage precision according to road usage frequency, ensuring perception effectiveness for nearby areas and controlling storage costs for distant areas, balancing cloud storage overhead with the perception precision of features retrieved by the vehicle, and solving the problem that uniform granular compression either results in distortion or excessive storage costs.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The application relates to the technical field of intelligent driving and discloses a vehicle-cloud cooperative intelligent driving method, a vehicle-cloud cooperative intelligent driving system and a medium, the method comprising the following steps: acquiring environment sensing data collected by a vehicle and extracting vehicle-side space features from the environment sensing data; sending current position information of the vehicle to a cloud end; receiving local space features issued by the cloud end in response to the current position information, wherein the local space features are feature vectors containing geometric information and semantic information and extracted by the cloud end based on original world perception data collected by a sensor; fusing the vehicle-side space features and the local space features to obtain global features; and inputting the global features into a vehicle control model so as to output a vehicle control instruction through the vehicle control model according to the global features; the application improves the response capability of vehicle-cloud cooperative intelligent vehicle control and reduces application cost.
Need to check novelty before this filing date? Find Prior Art