Lane selection system and method for autonomous vehicle

Abstract

The invention provides a lane selection method for autonomous driving. The processing steps are as follows: 1, the initial point, the ending point and the driving area of a road path are determined, and each road section and each lane are divided; 2, the vehicle speed of each lane is acquired, the average vehicle speed is calculated, and a real-time information database is built; 3, the average vehicle speed is sent to a vehicle-mounted terminal, and lane-level road path planning is carried out based on the shortest time; 4, when the changed real-time average vehicle speed enables the planned driving lane to be changed, according to the average speed difference between the lanes and the distance information between the vehicle position and a front intersection, whether to change a lane for driving is considered comprehensively; and 5, if lane changing can be carried out, the newly-planned lane is tracked for driving, and if not, the originally-planned lane is used for driving. Lane selection is carried out according to the real-time vehicle flow speed and the position information between the vehicle and the intersection, the driving lane is corrected automatically in real time, and the vehicle can be autonomously driven to avoid a congested road.

Description

technical field [0001] The invention belongs to the field of information fusion technology and intelligent information processing technology, and in particular relates to a lane selection method for an automatic driving vehicle. Background technique [0002] With the development of driver assistance technology, autonomous driving technology is increasingly becoming a topic of concern. Autonomous vehicles have higher requirements for path planning than traditional vehicles, and they need to be accurate to the lane level. At the same time, how to actively avoid congested road sections according to real-time traffic flow conditions has become a new problem faced by the autonomous vehicle path planning system. [0003] Patent Document 1 (CN104880193A) discloses a lane-level navigation system and its lane-level navigation method. According to vehicle positioning information and destination information, according to the current vehicle congestion index and the average driving spe...

AI Technical Summary

Problems solved by technology

[0006] For the system disclosed in Patent Document 1, manual intervention is required, and the driver chooses the best lane to avoid congestion, which lacks the ability of intelligent perception and intelligent path planning

The automatic driving system described in Patent Document 2 controls the safe driving of the vehicle through the positioning module and road images; the online path planning system for smart cars described in Patent Document 3 is an offline system, which only obtains road condition information around the vehicle through binocular cameras to control The vehicle is safe to drive. Patent Document 2 and Patent Document 3 cannot select an optimal lane for smart cars to avoid congestion within a long distance.

Images