Lane changing decision-making method and system based on lane changing interaction intention

Abstract

The invention discloses a lane changing decision-making method and system based on a lane changing interaction intention. The method comprises the following steps: collecting lane changing interaction data; marking the lane changing interaction data through road information and a lane scene, and extracting driving features from the lane changing interaction data to construct a training set and a verification set; labeling the driving features, and determining a transverse lane changing intention; training a long-short term memory network through the training set and the transverse lane changing intention to obtain a transverse lane changing intention model; training a long-short term memory network through the transverse lane changing intention and the driving features to obtain a longitudinal way giving intention model; according to the transverse lane changing intention and longitudinal way giving intention, judging the lane changing interaction state of a lane changing interaction vehicle; and determining a lane changing decision according to the lane changing interaction state. According to the method, the current lane changing interaction intention of the vehicle is identified, so that correct decision can be made when the automatic driving vehicle changes the lane in the lane changing scene or when surrounding vehicles change the lane, and the traffic safety in the lane changing process is improved.

Description

Technical field [0001] The present invention relates to the field of automatic driving techniques, and more particularly to a trifted decision making method and system based on a trail interaction interaction. Background technique [0002] With the increasing amount of the number of vehicles and the continuous improvement of the intelligence of the automatic driving vehicle, the understanding of the inter-vehicle interaction behavior has put forward higher requirements, and the study of the car interaction behavior is also more concerned. Change behavior is a common basic driving scenario, and the automatic driving vehicle trigger model has been widely studied. However, the study of vehicle interaction problems during the interior is not systematic. [0003] Vehicle violations, incoming driving operations such as transit and overtakes may cause traffic conflicts and collisions, and to a large extent affect the efficiency of traffic flow. In the rectangular scene, it is possible t...

AI Technical Summary

Problems solved by technology

Lane-changing behavior is a common basic driving scenario. The lane-changing model of autonomous driving vehicles has been extensively studied. However, the research on vehicle interaction in the process of lane-changing at home and abroad is not yet systematic enough.

[0003] Improper driving operations such as illegal lane changes, cut-in, and overtaking of vehicles may cause traffic conflicts and collisions, which greatly affect the efficiency of traffic flow

In the lane-changing scene, it may be difficult for the current self-driving vehicles to make reasonable traffic decisions due to high traffic density and many uncertain factors, and it is impossible to reasonably interact with surrounding traffic participants.

Images