CN-122024520-A - Collision early warning method and device, vehicle and storage medium

CN122024520ACN 122024520 ACN122024520 ACN 122024520ACN-122024520-A

Abstract

A collision early warning method is disclosed, and relates to the technical field of intelligent driving, and the method comprises the steps of obtaining an image of an environment where a vehicle is located, determining an identification result of the image, wherein the identification result comprises a traffic participant related to the vehicle, obtaining a first relative distance between the vehicle and the traffic participant, determining an initial collision prediction result according to the first relative distance and the identification result, performing space-time sequence prediction according to the identification result if the initial collision prediction result indicates that the traffic participant is a moving traffic participant, obtaining a first prediction track of the vehicle and a second prediction track of the traffic participant, and performing collision early warning on the vehicle if the vehicle is predicted to collide with the traffic participant according to the first prediction track and the second prediction track. The application can improve the precision and accuracy of collision early warning of the vehicle, and can provide prompt earlier before danger occurs, thereby obviously improving the safety and reliability of the vehicle in the high-speed running process.

Inventors

BAO SHUN
PAN NAN
HUANG CHEN
ZHA GUOZE
PAN XIAORUI
WANG XIANLI

Assignees

奇瑞汽车股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260120

Claims (10)

1. A collision warning method, the method comprising: Acquiring an image of an environment where a vehicle is located, and determining a recognition result of the image, wherein the recognition result comprises traffic participants related to the vehicle; Acquiring a first relative distance between the vehicle and the traffic participant, and truly initiating a collision prediction result according to the first relative distance and the identification result; if the initial collision prediction result indicates that the traffic participant is a moving traffic participant, performing space-time sequence prediction according to the identification result to obtain a second prediction track of the traffic participant; And acquiring a first predicted track of the vehicle, and predicting and determining that the vehicle collides with the traffic participant according to the first predicted track and the second predicted track, so as to perform collision early warning on the vehicle.
2. The method of claim 1, wherein the obtaining a first predicted trajectory of the vehicle and predicting that the vehicle will collide with the traffic participant based on the first predicted trajectory and the second predicted trajectory, pre-warning the vehicle for a collision, comprises: Determining a predicted collision point according to the first predicted track and the second predicted track; Acquiring a first current location of the vehicle and a second current location of the traffic participant; determining whether the vehicle collides with the traffic participant at the predicted collision point according to the first current position, the second current position and the predicted collision point; and if the fact that the vehicle and the traffic participant are about to collide at the predicted collision point is determined, collision early warning is carried out on the vehicle.
3. The method of claim 2, wherein the determining whether the vehicle collides with the traffic participant at the predicted point of collision based on the first current location, the second current location, and the predicted point of collision comprises: Determining a first duration of travel of the vehicle to the predicted collision point according to the first current position and the predicted collision point; Determining a second duration of travel of the traffic participant to the predicted collision point based on the second current location and the predicted collision point; If the first time period is longer or shorter than the second time period, determining that the vehicle and the traffic participant do not collide at the predicted collision point; And if the first time period is equal to the second time period, determining that the vehicle collides with the traffic participant at the predicted collision point.
4. A method according to claim 3, wherein said collision warning of said vehicle comprises: Acquiring a first speed of the vehicle and a second speed of the traffic participant, and determining a relative speed of the traffic participant compared to the vehicle based on the first speed and the second speed; determining a second relative distance from the first current location and the second current location; determining a target relative distance from the second relative distance, the relative velocity, and the first time period; Determining a collision prediction duration according to the target relative distance, the first speed and the second speed; If the predicted collision time is less than or equal to the preset time, generating early warning information, and carrying out collision early warning according to the early warning information.
5. The method of claim 4, wherein the collision pre-warning the vehicle further comprises: Determining a target category of the traffic participant according to the identification result, and determining a target weight of the traffic participant according to the target category; determining a risk score according to the relative speed, the collision prediction duration and the target weight; And determining a collision early warning strategy of the vehicle according to the risk score, and performing collision early warning on the vehicle according to the collision early warning strategy.
6. The method of claim 5, wherein the determining a collision warning strategy for the vehicle based on the risk information comprises: If the risk score is greater than or equal to a score threshold, determining that the collision early warning strategy is an intervention strategy; And if the risk score is smaller than the score threshold, determining the collision early warning strategy as an alarm strategy.
7. The method of any one of claims 1-6, wherein after the collision pre-warning of the vehicle, the method further comprises: Acquiring control information of the vehicle in a target duration; and if the control information indicates that the risk avoidance control does not exist in the target duration, generating an intervention control parameter, and performing collision intervention on the vehicle based on the intervention control parameter.
8. A collision warning device, the device comprising: The system comprises an identification module, a control module and a control module, wherein the identification module is used for acquiring an image of an environment where a vehicle is located and determining an identification result of the image, and the identification result comprises traffic participants related to the vehicle; The initial collision prediction result determining module is used for obtaining a first relative distance between the vehicle and the traffic participant and determining an initial collision prediction result according to the first relative distance and the identification result; The track prediction module is used for performing space-time sequence prediction according to the identification result to obtain a second predicted track of the traffic participant if the initial collision prediction result indicates that the traffic participant is a moving traffic participant; and the early warning module is used for acquiring a first predicted track of the vehicle, predicting and determining that the vehicle collides with the traffic participant according to the first predicted track and the second predicted track, and carrying out collision early warning on the vehicle.
9. A vehicle, characterized in that the vehicle comprises: A processor; a memory having stored thereon computer readable instructions which, when executed by the processor, implement the method of any of claims 1 to 7.
10. A computer readable storage medium having stored therein program code which is callable by a processor to perform the method of any one of claims 1 to 7.

Description

Collision early warning method and device, vehicle and storage medium Technical Field The application relates to the field of intelligent driving, in particular to a collision early warning method, a device, a vehicle and a storage medium. Background With the rapid development of technology, the application of intelligent driving functions in vehicles is also becoming more and more popular. Among the intelligent driving functions, predicting potential hazards in advance and performing early warning become one of the most important functions. At present, the early warning is usually realized by detecting the obstacle through the vehicle-mounted radar, but the vehicle-mounted radar has the advantages of small identification precision, long detection distance and incapability of adapting to various complex scenes. Therefore, how to improve the accuracy and precision of collision warning is a problem to be solved. Disclosure of Invention In view of the above, embodiments of the present application provide a collision warning method, device, vehicle and storage medium, so as to improve the above-mentioned problems. According to a first aspect of the embodiment of the application, a collision early warning method is provided, and the method comprises the steps of obtaining an image of an environment where a vehicle is located, determining a recognition result of the image, wherein the recognition result comprises a traffic participant related to the vehicle, obtaining a first relative distance between the vehicle and the traffic participant, determining an initial collision prediction result according to the first relative distance and the recognition result, performing space-time sequence prediction according to the recognition result if the initial collision prediction result indicates that the traffic participant is a moving traffic participant, obtaining a second prediction track of the traffic participant, obtaining the first prediction track of the vehicle, and performing collision early warning on the vehicle according to the first prediction track and the second prediction track. In some embodiments, the method comprises obtaining a first predicted track of the vehicle, predicting and determining that the vehicle will collide with the traffic participant according to the first predicted track and the second predicted track, and performing collision early warning on the vehicle, wherein the method comprises determining a predicted collision point according to the first predicted track and the second predicted track; the method comprises the steps of obtaining a first current position of a vehicle and a second current position of a traffic participant, determining whether the vehicle collides with the traffic participant at a predicted collision point according to the first current position, the second current position and the predicted collision point, and carrying out collision early warning on the vehicle if the vehicle is determined to collide with the traffic participant at the predicted collision point. In some embodiments, the determining whether the vehicle collides with the traffic participant at the predicted collision point according to the first current position, the second current position and the predicted collision point comprises determining a first duration of travel of the vehicle to the predicted collision point according to the first current position and the predicted collision point, determining a second duration of travel of the traffic participant to the predicted collision point according to the second current position and the predicted collision point, determining that the vehicle does not collide with the traffic participant at the predicted collision point if the first duration is greater than or less than the second duration, and determining that the vehicle collides with the traffic participant at the predicted collision point if the first duration is equal to the second duration. In some embodiments, the collision early warning of the vehicle comprises the steps of obtaining a first speed of the vehicle and a second speed of the traffic participant, determining the relative speed of the traffic participant compared with the vehicle according to the first speed and the second speed, determining a second relative distance according to the first current position and the second current position, determining a target relative distance according to the second relative distance, the relative speed and the first time, determining a collision prediction time according to the target relative distance, the first speed and the second speed, generating early warning information if the predicted collision time is smaller than or equal to a preset time, and performing collision early warning according to the early warning information. In some embodiments, the method for collision early warning of the vehicle further comprises determining a target class of the traffic participant according to