CN-116501821-B - Vehicle and method and device for acquiring and automatically driving functional difference layers of vehicle

CN116501821BCN 116501821 BCN116501821 BCN 116501821BCN-116501821-B

Abstract

The embodiment of the application discloses a vehicle and a method and a device for acquiring and automatically driving a functional difference layer of the vehicle, wherein the method for acquiring the functional difference layer comprises the steps of starting an automatic driving function of a road test vehicle on a real road with high-precision map coverage; the method comprises the steps of obtaining running data of a road test vehicle on a real road, wherein the running data comprise first abnormal point data marked when automatic driving functions are abnormal, restoring a running process on a constructed electronic horizon road model according to the running data, correspondingly marking second abnormal point data according to the first abnormal point data in the restored running process, obtaining corresponding model data, wherein the model data comprise the second abnormal point data, and generating a functional difference layer corresponding to the current road scene type according to the model data. By the scheme of the embodiment, the automatic driving function of the vehicle is helped to exit or degrade in time, and driving safety is ensured.

Inventors

LI QIAO
FU LIANG
HUANG HU
ZHANG XIAOLEI
FU JIE
WANG WEI

Assignees

宁波吉利汽车研究开发有限公司
浙江吉利控股集团有限公司

Dates

Publication Date: 20260512
Application Date: 20230424

Claims (11)

1. The method for acquiring the functional difference layer is characterized by comprising the following steps: Starting an automatic driving function of the road test vehicle on a real road with map coverage; Acquiring running data of the road test vehicle on the real road, wherein the running data comprises first abnormal point data marked when the automatic driving function is abnormal; restoring a running process on the constructed electronic horizon road model according to the running data, correspondingly marking second abnormal point data according to the first abnormal point data in the restored running process, and acquiring corresponding model data; Generating a functional difference layer corresponding to the current road scene type according to the model data, wherein the functional difference layer comprises a blacklist or a whitelist of the driving road sections of the real road generated according to the second abnormal point data; the blacklist comprises road section information corresponding to the abnormal running road section when the abnormal running road section is less than the normal running road section on the real road scene; The white list comprises road section information corresponding to a normally-represented running road section when the normally-represented running road section is more than the normally-represented running road section on a real road scene; the road section information comprises road identification ID, lane ID and/or lane number data.
2. The method for acquiring the functional difference layer according to claim 1, wherein the acquiring the driving data of the road-test vehicle on the real road includes: Acquiring road information in the process of driving the road test vehicle on the real road, and marking the moment and the position of the abnormal automatic driving function as the first abnormal point data; and capturing electronic horizon message data containing the road information and the first abnormal point data as the driving data.
3. The method for obtaining a functional difference layer according to claim 2, wherein the restoring the driving process on the constructed electronic horizon road model according to the driving data includes: loading the electronic horizon message data by using a preset map visualization tool, constructing the electronic horizon road model, and displaying the electronic horizon road model in the map visualization tool as a base map; and playing back the electronic horizon message data on the electronic horizon road model, displaying the positions of the road test vehicle on the real road at different moments in a simulation mode on the base map of the map visualization tool, and displaying each item of road information in the electronic horizon message data.
4. The method for acquiring a functional difference layer according to claim 3, wherein the first outlier data and the second outlier data each include a time and a position when the autopilot is dysfunctional; And marking second abnormal point data correspondingly in the restored driving process according to the first abnormal point data, wherein the method comprises the following steps: And comparing the first abnormal point data, marking the second abnormal point data in the map visualization tool by combining the real-time position of the road test vehicle which is replayed in the map visualization tool, and automatically writing the position attribute value related to the second abnormal point data into a mark file in a preset format.
5. The method for obtaining a functional difference layer according to claim 4, wherein the generating the functional difference layer corresponding to the current road scene type according to the model data includes: Reading the annotation file, classifying the annotation file into the current road scene type, generating a functional difference layer file corresponding to the current road scene type by taking a city as a unit, and acquiring the functional difference layer based on the functional difference layer file.
6. The method of claim 5, wherein each of the road scene types includes a road scene switch; the road scene switch is arranged to turn on the road scene and control the electronic horizon road module in the road scene to send the electronic horizon message data when in an on state, and turn off the road scene and control the electronic horizon road module in the road scene not to send the electronic horizon message data when in an off state.
7. A functional difference layer acquisition device comprising a first processor and a first memory, wherein a first instruction is stored in the first memory, and wherein the method for acquiring a functional difference layer according to any one of claims 1-6 is implemented when the first instruction is executed by the first processor.
8. An automated driving method, the method comprising: Invoking a functional difference layer corresponding to a current driving road, wherein the functional difference layer is acquired according to the method for acquiring the functional difference layer according to any one of claims 1-6; and carrying out automatic driving according to the called functional difference layer.
9. The method of autopilot of claim 8 wherein said automatically driving in accordance with said retrieved functional difference layer comprises: determining the road scene type corresponding to the functional difference layer; When the determined driving road sections with abnormal performance in the road scene type are less than the driving road sections with normal performance, setting a preset road scene switch in an on state, marking a black-and-white list mark corresponding to the road scene type as a black list, so that the vehicle can execute an automatic driving function on the driving road sections except the black list on the current driving road, and stopping the automatic driving function on the driving road sections corresponding to the black list; When the abnormal driving road sections are more than the normal driving road sections in the determined road scene type, setting the road scene switch in an off state, marking the black-and-white list mark as a white list, enabling the driving road sections of the vehicle on the current driving road except the white list to stop an automatic driving function, and executing the automatic driving function on the driving road sections corresponding to the white list.
10. An autopilot comprising a second processor and a second memory, the second memory having stored therein second instructions, wherein the autopilot method of claim 8 or 9 is implemented when the second instructions are executed by the second processor.
11. A vehicle characterized by comprising the functional difference layer acquisition device according to claim 7 and the automatic driving device according to claim 10.

Description

Vehicle and method and device for acquiring and automatically driving functional difference layers of vehicle Technical Field The embodiment of the application relates to a vehicle control technology, in particular to a vehicle and a method and a device for acquiring and automatically driving a functional difference layer of the vehicle. Background With the continuous development of technology, intelligent driving of automobiles becomes an important research and development direction of various large-vehicle enterprises. While each large vehicle enterprise seeks a higher level (l3+) of intelligent driving function, intelligent driving safety is also receiving more and more attention from each large vehicle enterprise. Under ideal conditions, the high-level (L3+) intelligent driving function can well ensure the driving safety of the automobile under the addition of a plurality of perception sensors and software. In reality, however, in complex road and traffic scenarios, the intelligent driving function often cannot guarantee the driving safety in a percentage. Meanwhile, the intelligent driving function is limited by the objective development level of the current intelligent technology, and can not safely process all roads and traffic scenes, so that corresponding driving safety problems can be caused. Disclosure of Invention The embodiment of the application provides a vehicle and a method and a device for acquiring and automatically driving a functional difference layer thereof, which can help the automatic driving function of the vehicle to exit or degrade in time and ensure the driving safety. The embodiment of the application provides a method for acquiring a functional difference layer, which can comprise the following steps: Starting an automatic driving function of the road test vehicle on a real road covered by a high-precision map; Acquiring running data of the road test vehicle on the real road, wherein the running data comprises first abnormal point data marked when the automatic driving function is abnormal; restoring a running process on the constructed electronic horizon road model according to the running data, correspondingly marking second abnormal point data according to the first abnormal point data in the restored running process, and acquiring corresponding model data; And generating a functional difference layer corresponding to the current road scene type according to the model data, wherein the functional difference layer comprises a blacklist or a whitelist of the driving road sections of the real road, which are generated according to the second abnormal point data. In an exemplary embodiment of the present application, the acquiring the driving data of the road test vehicle on the real road may include: Acquiring road information in the process of driving the road test vehicle on the real road, and marking the moment and the position of the abnormal automatic driving function as the first abnormal point data; and capturing electronic horizon message data containing the road information and the first abnormal point data as the driving data. In an exemplary embodiment of the present application, the restoring the driving process on the constructed electronic horizon road model according to the driving data may include: loading the electronic horizon message data by using a preset map visualization tool, constructing the electronic horizon road model, and displaying the electronic horizon road model in the map visualization tool as a base map; and playing back the electronic horizon message data on the electronic horizon road model, displaying the positions of the road test vehicle on the real road at different moments in a simulation mode on the base map of the map visualization tool, and displaying each item of road information in the electronic horizon message data. In an exemplary embodiment of the present application, the first outlier data and the second outlier data may each include a time and a location when the autopilot function is abnormal; the marking the second outlier data according to the first outlier data in the restored driving process may include: And comparing the first abnormal point data, marking the second abnormal point data in the map visualization tool by combining the real-time position of the road test vehicle which is replayed in the map visualization tool, and automatically writing the position attribute value related to the second abnormal point data into a mark file in a preset format. In an exemplary embodiment of the present application, the generating a functional difference layer corresponding to a current road scene type according to the model data may include: Reading the annotation file, classifying the annotation file into the current road scene type, generating a functional difference layer file corresponding to the current road scene type by taking a city as a unit, and acquiring the functional difference layer based on the functional differ