5G C-V2X vehicle road cloud cooperation perception method and system used for automatic driving vehicle

Abstract

The invention provides a 5G C-V2X vehicle road cloud cooperation perception method and system used for an automatic driving vehicle. Rapid acquisition and transmission of various information on a roadcan be realized through a 5G technology and a vehicle networking technology, and a dynamic scene is constructed according to collected information, a running path of a target vehicle in the dynamic scene is simulated, and an optimal path is selected so that the target vehicle can achieve automatic driving with the aid of the optimal path. The collected information can include information of a roadside obstacle, traffic light, a pedestrian, a running vehicle and the like, and data sharing is performed through the 5G technology so as to provide sufficient information for the automatic driving and ensure rapidity and correctness of an automatic driving vehicle decision.

Description

technical field [0001] The present invention relates to the technical field of automatic driving, in particular to a 5G C-V2X vehicle-road-cloud cooperative sensing method and system for automatic driving vehicles. Background technique [0002] V2X (vehicle to everything) Internet of Vehicles can realize the communication between vehicles and vehicles, vehicles and people, vehicles and roadsides, and vehicles and cloud platforms, realize the sharing of information, and make autonomous driving bid farewell to the era of "single-vehicle intelligence". [0003] At present, typical self-driving smart car systems generally use on-board sensors to obtain information such as road scenes and vehicle attitudes, and then determine vehicle body control commands (such as power, direction, brakes, etc.) through path planning and driving behavior decision-making mechanisms. In the process, the real-time information obtained by the sensor is combined with the feedback mechanism to realize ...

AI Technical Summary

Problems solved by technology

[0003] At present, typical self-driving smart car systems generally use on-board sensors to obtain information such as road scenes and vehicle attitudes, and then determine vehicle body control commands (such as power, direction, brakes, etc.) through path planning and driving behavior decision-making mechanisms. In the process, the real-time information obtained by the sensor is combined with the feedback mechanism to realize adaptive driving control and driving safety. The use of on-board sensors for information acquisition can certainly obtain unknown information around the vehicle and carry out target recognition, classification and processing, but only relying on the single Vehicle sensors have limited perception capabilities and processing speeds, and each sensor has its own shortcomings. No matter what kind of sensor or fusion perception of current self-driving vehicles, it is impossible to overcome obstacles and space-time constraints to realize the surrounding environment. Comprehensive perception cannot solve the problem of visual occlusion and driving safety in blind spots, etc., which constitute the bottleneck problem restricting the development of autonomous vehicles

Images