Design And Implementation Of Cloud Data Platform For Intelligent Connected Vehicles

With the development of autonomous driving technology,the intelligence level of intelligent vehicle has become higher and higher.However,in the process of improving the level of autonomous driving of intelligent vehicle,the improvement of the intelligence of single intelligent vehicle cannot effectively solve the functions that need to be realized in the process of high-level autonomous driving,such as global intelligent traffic planning and scheduling,beyond visual range perception and so on.The combination of intelligent vehicle and cloud computing technology to develop cloud control platform is a feasible technical scheme for the development of advanced autonomous driving technology.Therefore,this paper has carried out in-depth research and exploration on the task of developing the intelligent connected vehicles cloud platform.Based on the key project of laboratory,this paper conducts research on the quick landing scheme of small standardized intelligent connected vehicles edge cloud platform,and carries out relevant system development and experimental verification based on the current specific requirements of intelligent connected vehicles and the construction scheme of cloud platform.The main work contents are as follows:1.Firstly,the paper analyzes the development trend of intelligent vehicle in networking and the application status of intelligent connected vehicles cloud platform.Aiming at the low delay and high security requirements of intelligent networked vehicles,this paper proposes an intelligent connected vehicles edge cloud platform architecture,and on this basis,implements the corresponding landing scheme.2.On this basis,according to the functional requirements of edge cloud platform in the scene of Internet of vehicles,there are few standardized solutions based on edge cloud in the field of Internet of vehicles.In this paper,we hope to start with the underlying technology of the cloud platform to realize an edge cloud landing scheme that can be flexibly deployed.Using mainstream open source hardware virtualization software Docker and Kubernetes cluster management engine lab server resources to container and effective integration,and on the basis of the deployment of the basis for subsequent development used container software,for the realization of the function of cloud data platform specific to provide the corresponding support,eventually build up edge cloud cloud PaaS layer foundation platform.3.Based on the functional requirements of cloud,storage and visualization of intelligent vehicle-side data in real time,the three subsystems of vehicle cloud communication,cloud storage and cloud data visualization are built in the cloud foundation platform to realize the SaaS functional application layer of edge cloud.Aiming at the problem that the real-time video transmission and display cannot be realized due to the insufficient network bandwidth,the edge cloud scheme of this paper adopts the two-channel vehicle-cloud data transmission scheme to send back the message data and image data of vehicle-end in real time respectively.The experimental results show that the frame rate of cloud real-time video display can reach more than 5 frames /s.Compared with the central cloud vehicle network scheme,the edge cloud scheme proposed in this paper has a great improvement in real-time performance.The intelligent connected vehicles cloud data platform designed and realized in this paper is a feasible landing scheme for the proposed intelligent connected vehicles edge cloud platform architecture.Different from the internet-of-vehicles solution based on the central cloud,this solution pays more attention to building a low-latency and highly reliable edge cloud platform between the smart car and the central cloud,which provides a feasible solution for the landing of internet-of-vehicle technology in the edge scenario.