Abstract:

A design for a C/S (Client/Server) model-based real-time multidimensional information interaction system was presented to enhance the safety and operational efficiency of autonomous vehicles in low-speed, semi-closed scenarios, utilizing the Robot Operating System 2 (ROS2) framework. The research incorporated data transmission fusion technology, which combined various data preprocessing algorithms with a unified network transmission protocol, enabling efficient transmission and real-time interaction of images, point clouds, and vehicle status information. The results show that under excellent network conditions, the average delay of the system is less than 0.063 seconds, the receiving frequency is almost equal to the sending frequency, and the packet loss rate does not exceed 3.5%; The integrated remote control module ensures the accuracy of operations; Under long-term stable operation, the system still maintains a low central processing unit (CPU) usage rate of 2.7%, demonstrating excellent resource efficiency. The system has demonstrated effectiveness in ensuring real-time, reliability, and low resource consumption, and can provide strong support for information exchange in the field of autonomous driving.

Key words: autonomous driving, semi enclosed scenes, multidimensional information exchange, real time performance, data transmission fusion technology, cloud control