A 5G-V2X and vehicle-road-cloud collaborative system for intelligent connected vehicles: architecture, algorithm, and experimental verification
TL;DR
A hierarchical architecture model and fusion perception algorithm based on 5G-V2X, and a multivariate data fusion method combining Kalman filter and attention mechanism neural network are proposed, which significantly improves the environmental perception and collaborative decision-making ability of vehicles.
Abstract
With the acceleration of the global digitalization process, intelligent connected vehicles have become the strategic direction for the transformation and upgrading of the automobile industry. This paper focuses on the application of 5G technology in the innovation of intelligent connected automobile industry to carry out systematic research, builds a hierarchical architecture model and fusion perception algorithm based on 5G-V2X, and proposes a multivariate data fusion method combining Kalman filter and attention mechanism neural network, which significantly improves the environmental perception and collaborative decision-making ability of vehicles. Through multi-scenario experimental verification, it can be seen that the system can reduce the communication delay to 8.2 milliseconds, the positioning accuracy to 12.3 cm, the emergency braking frequency to 65.71% and the traffic efficiency to 19.06% in the 5G network environment, which effectively solves the core problems of intelligent connected vehicles in real-time, safety and traffic collaborative optimization. To sum up, the research results clarify the key roles of 5G-V2X communication, edge computing and vehicle-road-cloud collaboration mechanism in the integration of production and research and commercialization, and provide theoretical support and practical reference for the industrial innovation path.