Risk assessment algorithm introducing target behavior and collision probability

Abstract

The invention discloses a risk assessment algorithm introducing a target behavior and a collision probability, which is characterized in that a vehicle coordinate system is taken as a reference, a passing domain of a vehicle is planned according to the vehicle speed and a target relative distance, and through target filtering of the passing domain, a target which is in the passing domain or is predicted to enter the passing domain is considered; the method comprises the following steps: performing behavior category judgment and collision probability calculation, when a target behavior category is Run-forward and the collision probability is greater than a threshold value, considering that a target has a collision risk, outputting a collision target for decision and return control algorithm reference, and finally judging whether braking needs to be triggered in combination with parameters such as TTC, TTB, required minimum acceleration and the like. According to the invention, false triggering and missed triggering can be better reduced, and the driving safety is improved.

Description

technical field [0001] The present invention relates to an algorithm related to the active safety function of automobile automatic driving, and specifically relates to a method of processing and predicting the distance, speed, and acceleration information of the target ahead, combined with the attitude state of the vehicle and the behavior of the driver, to determine whether the target will Entering or already in the self-vehicle traffic area, combined with the prediction algorithm to identify the behavior category and collision probability of the target in front, confirm whether there is a risk of collision, so as to further assist in solving the problem of false triggering or missed triggering of safety assistance functions. Background technique [0002] During the automatic driving process of the car, the road environment is complex and the target attributes are different: there may be target types such as vehicles, pedestrians, bicycles, two-wheeled vehicles, obstacles, a...

AI Technical Summary

Problems solved by technology

The safety control module of autonomous driving needs to accurately identify potentially dangerous targets from environmental perception, so as to avoid the risk of collision, which poses a great challenge to the algorithm

Images