Automatic driving vehicle path planning system and method based on V2I technology

Abstract

The invention discloses an automatic driving vehicle path planning system and method based on a V2I technology, and relates to the technical field of automatic driving. According to the invention, a vehicle-mounted sensor of an automatic driving vehicle, an electronic high-definition map system and the V2I technology are combined; the vehicle position is mapped to MAP data broadcasted by RSU equipment on a roadside by utilizing data acquired by the vehicle-mounted sensor; related information of a signal lamp phase which should be concerned by a current lane and a connection mode of the currentlane in a road network can be obtained by combining SPAT data broadcasted by RSU equipment; therefore, a basis is provided for path planning and speed planning of the automatic driving vehicle, accurate driving of the automatic driving vehicle can be achieved, the electronic high-definition map system and MAP data are tightly combined, and the problem that the electronic high-definition map system is insufficient in accuracy is solved.

Description

technical field [0001] The invention relates to the technical field of automatic driving, in particular to a V2I technology-based automatic driving vehicle path planning system and method. Background technique [0002] The path planning of autonomous driving vehicles is similar to the navigation system of traditional vehicles in function. It mainly provides direction instructions for vehicles driving in the road network based on high-precision maps. The vehicle provides accurate lane-level guidance routes, signal light phase information and speed limit information, etc., so that the vehicle can automatically make corresponding decisions. [0003] Although the current high-precision map can provide high-precision lane-level positioning for self-driving vehicles, it cannot provide dynamic information on signal lights and the connection mode of the traffic network at intersections, so it cannot provide guidance for the speed and path planning of self-driving vehicles. . Curre...

AI Technical Summary

Problems solved by technology

[0003] Although the current high-precision map can provide high-precision lane-level positioning for self-driving vehicles, it cannot provide dynamic information on signal lights and the connection mode of the traffic network at intersections, so it cannot provide guidance for the speed and path planning of self-driving vehicles.

Currently, self-driving vehicles usually obtain images of signal lights at intersections through on-board cameras and use image recognition to judge the phase information of signal lights. However, this method is seriously affected by the environment and cannot obtain accurate countdown information, so it is not accurate.

Images