CN107221195B - Automobile lane prediction method and lane level map - Google Patents
Automobile lane prediction method and lane level map Download PDFInfo
- Publication number
- CN107221195B CN107221195B CN201710384424.6A CN201710384424A CN107221195B CN 107221195 B CN107221195 B CN 107221195B CN 201710384424 A CN201710384424 A CN 201710384424A CN 107221195 B CN107221195 B CN 107221195B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- road section
- road
- dimensional map
- lane
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/16—Anti-collision systems
- G08G1/166—Anti-collision systems for active traffic, e.g. moving vehicles, pedestrians, bikes
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/16—Anti-collision systems
- G08G1/167—Driving aids for lane monitoring, lane changing, e.g. blind spot detection
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses an automobile lane prediction method and a lane level map, which can judge the level of a current driving road section of a vehicle by using historical track data and lane level map data, particularly judge whether the vehicle and an object vehicle are in the same road section level and judge whether collision danger exists in a three-dimensional space range or not under the condition of layering at a road space level conversion part such as an overpass, and further inhibit false alarm.
Description
Technical Field
The invention belongs to the field of intelligent driving of vehicles, and particularly relates to a road self-adaptive prediction method.
Background
An Intersection Collision Warning System (ICWS) is an important component of an Advanced Driver Assistance System (ADAS), and can prevent Intersection Collision accidents and increase Intersection traffic efficiency for road traffic.
In the prior art, an intersection collision early warning system based on radar and a V2X system can only be used for two-dimensional plane traffic, vehicles in different plane levels are easily mistakenly taken as potential threats in three-dimensional traffic to generate false early warning, and normal driving of a driver is influenced.
Therefore, it is necessary to provide a vehicle road prediction method based on a three-dimensional space, which is to preferentially determine spatial road information of a host vehicle and a target vehicle before vehicle collision warning, so as to avoid false collision warning of vehicles in a road segment where the space does not intersect.
Disclosure of Invention
According to the automobile lane prediction method disclosed by the invention, the high-frequency road section ID is mapped through the historical track point of the automobile and the two-dimensional map, so that the road section ID number of the current automobile is accurately judged, and the prediction accuracy of the future driving path of the automobile is improved.
The invention discloses a method for predicting a vehicle lane, which comprises the following steps,
1) calling a lane-level map in the vehicle information, wherein the lane-level map comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space layers of each road on the two-dimensional map;
2) collecting and storing historical tracks:
acquiring a group of vehicle position and course information data in a period of time before the current time point through satellite positioning to form a group of vehicle driving historical track points;
3) judging the ID number of the road section where the current vehicle is located:
projecting a group of vehicle driving historical track points in the step 2) onto a two-dimensional map in the lane level map, wherein one vehicle driving historical track point corresponds to a plurality of road section ID numbers, obtaining road section IDs with high frequency of occurrence of the road section ID numbers corresponding to the group of vehicle driving historical track points, and setting the road section ID with the highest frequency of occurrence as the road section ID number in the three-dimensional space where the current vehicle is located.
Further, the method also comprises the following steps,
4) predicting future driving track and alarming
Predicting the future driving track of the vehicle on the two-dimensional map according to the vehicle driving historical track points and the vehicle yaw velocity of the vehicle, or judging the future driving track of the vehicle according to the ID number of the road section where the current vehicle is located and the current vehicle position on the two-dimensional map;
obtaining future driving tracks of surrounding object vehicles through vehicle communication;
when the object vehicle on the two-dimensional map is found to drive into the cross road section and has collision risk, judging whether the vehicle intersects with the path of the object vehicle in the three-dimensional space according to the ID number of the road section where the current vehicle is located and the ID number of the road section where the object vehicle is located;
intersecting and alarming; and the alarm is not given when the alarm is not intersected.
The invention also discloses an automobile lane prediction method, which maps the high-frequency road section ID through the historical track point of the automobile and the two-dimensional map, obtains the road shape matching road section ID through matching the historical track shape of the automobile and the shapes of a plurality of paths on the two-dimensional map corresponding to the current position of the automobile, and fuses the high-frequency road section ID and the road shape matching road section ID, thereby accurately judging the serial number of the road section where the current automobile is located and improving the prediction accuracy of the future driving path of the automobile.
The invention discloses a method for predicting a vehicle lane, which comprises the following steps,
1) calling a lane-level map in the vehicle information, wherein the lane-level map comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space layers of each road on the two-dimensional map;
2) collecting and storing historical tracks:
acquiring position and course information data of a group of vehicles in a period of time before the current time point through satellite positioning to form a group of historical track points of vehicle running;
3) judging the ID number of the road section where the current vehicle is located:
projecting a group of vehicle driving historical track points in the step 2) onto a two-dimensional map in the lane level map, wherein one vehicle driving historical track point corresponds to a plurality of road section ID numbers, and obtaining road section IDs with high frequency of occurrence of the road section ID numbers corresponding to the group of vehicle driving historical track points;
matching a track shape formed by historical track points of vehicle driving with road shapes of a plurality of road IDs in a two-dimensional map corresponding to the current vehicle position to obtain road shape matching road section ID numbers;
and (4) fusing and judging the ID number of the road section with the shape matching road section and the ID number of the road section with high occurrence frequency to obtain the ID number of the road section where the current vehicle is located.
Further, the road shape matching road section ID number and the road section ID number with high occurrence frequency are fused and judged to be a weighted fusion method, and the corresponding road section ID number with high matching degree and high occurrence frequency is selected as the road section ID number where the vehicle is located.
Further, the method also comprises the following steps,
4) predicting future driving track and alarming
Predicting the future driving track of the vehicle on the two-dimensional map according to the vehicle driving historical track points and the vehicle yaw velocity of the vehicle, or judging the future driving track of the vehicle according to the ID number of the road section where the current vehicle is located and the current vehicle position on the two-dimensional map;
obtaining future driving tracks of surrounding object vehicles through vehicle communication;
when the object vehicle on the two-dimensional map is found to drive into the cross road section and has collision risk, judging whether the vehicle intersects with the path of the object vehicle in the three-dimensional space according to the ID number of the road section where the current vehicle is located and the ID number of the road section where the object vehicle is located;
intersecting and alarming; and the alarm is not given when the alarm is not intersected.
The invention also discloses an automobile lane prediction method, which maps the high-frequency road section ID through the historical track point of the automobile and the two-dimensional map, obtains the road direction matching ID number through the road direction matching of a plurality of road IDs in the two-dimensional map corresponding to the current course and the current position of the automobile, and fuses the high-frequency road section ID and the road direction matching road section ID, thereby accurately judging the road section ID number of the current automobile and improving the prediction accuracy of the future driving path of the automobile.
The invention discloses a method for predicting a vehicle lane, which comprises the following steps,
1) calling a lane-level map in the vehicle information, wherein the lane-level map comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space layers of each road on the two-dimensional map;
2) collecting and storing historical tracks:
acquiring position and course information data of a group of vehicles in a period of time before the current time point through satellite positioning to form a group of historical track points of vehicle running;
3) judging the ID number of the road section where the current vehicle is located:
projecting a group of vehicle driving historical track points in the step 2) onto a two-dimensional map in the lane level map, wherein one vehicle driving historical track point corresponds to a plurality of road section ID numbers, and obtaining road section IDs with high frequency of occurrence of the road section ID numbers corresponding to the group of vehicle driving historical track points;
matching the current course of the vehicle with the road directions of a plurality of road IDs in a two-dimensional map corresponding to the current vehicle position to obtain road direction matching ID numbers;
and (4) fusing and judging the road direction matching ID number and the road section ID number with high occurrence frequency to obtain the road section ID number of the current vehicle.
Further, the road direction matching road section ID number and the road section ID number with high occurrence frequency are fused and judged to be a weighted fusion method, and the corresponding road section ID number with high matching degree and high occurrence frequency is selected as the road section ID number where the vehicle is located.
Further, the method also comprises the following steps,
4) predicting future driving track and alarming
Predicting the future driving track of the vehicle on the two-dimensional map according to the vehicle driving historical track points and the vehicle yaw velocity of the vehicle, or judging the future driving track of the vehicle according to the ID number of the road section where the current vehicle is located and the current vehicle position on the two-dimensional map;
obtaining future driving tracks of surrounding object vehicles through vehicle communication;
when the object vehicle on the two-dimensional map is found to drive into the cross road section and has collision risk, judging whether the vehicle intersects with the path of the object vehicle in the three-dimensional space according to the ID number of the road section where the current vehicle is located and the ID number of the road section where the object vehicle is located;
intersecting and alarming; and the alarm is not given when the alarm is not intersected.
The invention also discloses a vehicle lane prediction method, which comprises the steps of mapping a high-frequency road section ID through a vehicle historical track point and a two-dimensional map, matching a plurality of path shapes on the two-dimensional map corresponding to the current position of the vehicle through the vehicle historical track shape to obtain a road shape matching road section ID, matching the current course of the vehicle with the road directions of the plurality of road IDs in the two-dimensional map corresponding to the current position of the vehicle to obtain a road direction matching ID number, and fusing the high-frequency road section ID, the road shape matching road section ID and the road direction matching road section ID, so that the road section ID number where the current vehicle is located is accurately judged, and the prediction accuracy of the future driving path of the vehicle is improved.
The invention discloses a method for predicting a vehicle lane, which comprises the following steps,
1) calling a lane-level map in the vehicle information, wherein the lane-level map comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space layers of each road on the two-dimensional map;
2) collecting and storing historical tracks:
acquiring position and course information data of a group of vehicles in a period of time before the current time point through satellite positioning to form a group of historical track points of vehicle running;
3) judging the ID number of the road section where the current vehicle is located:
projecting a group of vehicle driving historical track points in the step 2) onto a two-dimensional map in the lane level map, wherein one vehicle driving historical track point corresponds to a plurality of road section ID numbers, and obtaining road section IDs with high frequency of occurrence of the road section ID numbers corresponding to the group of vehicle driving historical track points;
matching a track shape formed by historical track points of vehicle driving with road shapes of a plurality of road IDs in a two-dimensional map corresponding to the current vehicle position to obtain road shape matching road section ID numbers;
matching the current course of the vehicle with the road directions of a plurality of road IDs in a two-dimensional map corresponding to the current vehicle position to obtain road direction matching ID numbers;
and (4) merging and judging the road section ID number with high frequency, the road shape matching road section ID number and the road direction matching ID number to obtain the road section ID number of the current vehicle.
Furthermore, the road section ID number with high frequency of occurrence, the road shape matching road section ID number and the road direction matching ID number are fused and judged as a weighted fusion method, and the corresponding road section ID number with high shape and direction matching degree and high frequency of occurrence is selected as the road section ID number where the vehicle is located.
Further, the method also comprises the following steps,
4) predicting future driving track and alarming
Predicting the future driving track of the vehicle on the two-dimensional map according to the vehicle driving historical track points and the vehicle yaw velocity of the vehicle, or judging the future driving track of the vehicle according to the ID number of the road section where the current vehicle is located and the current vehicle position on the two-dimensional map;
obtaining future driving tracks of surrounding object vehicles through vehicle communication;
when the object vehicle on the two-dimensional map is found to drive into the cross road section and has collision risk, judging whether the vehicle intersects with the path of the object vehicle in the three-dimensional space according to the ID number of the road section where the current vehicle is located and the ID number of the road section where the object vehicle is located;
intersecting and alarming; and the alarm is not given when the alarm is not intersected.
The invention also discloses a lane-level map, which comprises a two-dimensional map and a unique road section ID number capable of distinguishing the three-dimensional space hierarchy of each road on the two-dimensional map.
Further, the lane-level map is saved on the vehicle or on a cloud server.
The beneficial technical effects of the invention are as follows:
1) the lane level map comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space levels of each road on the two-dimensional map, is different from a common three-dimensional map library, and the three-dimensional map library distinguishes different space road sections through unique ID marks, can compress three-dimensional data to the maximum extent, and simultaneously meets the requirement of acquiring three-dimensional data of the road sections to judge the driving of the road sections.
2) The method comprises the steps of obtaining historical track points of vehicle running, mapping each track point to a two-dimensional map, enabling each track point to correspond to a plurality of paths, uniquely judging the ID of a three-dimensional space road section where a current vehicle is located by counting the ID number of a road section with high frequency in the historical track points, and improving the accuracy of judging the path where the current position is located by a method that the historical track corresponds to the path probability of the two-dimensional map, wherein the method is high in calculation efficiency.
3) The shapes corresponding to the vehicle driving historical track points are matched with the shapes of a plurality of road sections mapped on the two-dimensional map by the current position of the vehicle, the current road section ID with high matching degree can be matched, the position judgment of a three-dimensional space can be realized by utilizing two-dimensional data, and the judgment efficiency is improved;
4) matching the current course of the vehicle with the courses of a plurality of road sections on a two-dimensional map corresponding to the current position of the vehicle, matching the current road section ID with high matching degree, and utilizing two-dimensional data to realize position judgment of a three-dimensional space by utilizing the two-dimensional data, thereby improving the judgment efficiency;
5) and (4) carrying out weighted fusion on the three groups of road section ID judgment results with the advantages of 2, 3 and 4, and improving the judgment precision of the current road section ID.
6) Judging the collision possibility of the vehicle and the object vehicle on the two-dimensional map, judging whether the IDs of the three-dimensional space road sections are intersected and not intersected, and not giving an alarm when the collision danger exists; greatly reduced wrong report, solve the wrong report problem of judging vehicle collision based on two-dimensional space among the prior art.
Drawings
FIG. 1 is a block diagram of a vehicle system according to the present invention;
FIG. 2 is a first flowchart of a method for determining vehicle collision using the present vehicle lane prediction method;
FIG. 3 is a second flowchart for determining a vehicle collision using the present vehicle lane prediction method;
FIG. 4 is a third flowchart for determining vehicle collisions using the present vehicle lane prediction method;
FIG. 5 is a fourth flowchart for determining vehicle collisions using the present vehicle lane prediction method;
FIG. 6 is an application scenario diagram of the present vehicle lane prediction method for determining vehicle collision;
the system comprises a 1-short-range wireless communication module, a 2-satellite positioning module, a 3-lane map library module, a 4-control unit module, a 5-alarm module, a 6-vehicle and a 7-object vehicle.
Detailed Description
The invention will be further explained with reference to the drawings.
As shown in fig. 2, the lane prediction method for a vehicle, includes the steps of,
1) calling a lane-level map in vehicle information, wherein the lane-level map comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space layers of each road on the two-dimensional map;
2) collecting and storing historical tracks:
acquiring a group of vehicle position and course information data in a period of time before the current time point through satellite positioning to form a group of vehicle driving historical track points;
3) judging the ID number of the road section where the current vehicle is located:
projecting a group of vehicle driving historical track points in the step 2) onto a two-dimensional map in the lane level map, wherein one vehicle driving historical track point corresponds to a plurality of road section ID numbers, obtaining road section IDs with high frequency of occurrence of the road section ID numbers corresponding to the group of vehicle driving historical track points, and setting the road section ID with the highest frequency of occurrence as the road section ID number in the three-dimensional space where the current vehicle is located.
Therefore, the lane prediction is completed and can be used as a basis for judging the driving of the intelligent automobile, for example, the following step 4) is added for judging the collision. The basic application principle is as follows: when the ID number of the road section in the three-dimensional space where the current vehicle is located is determined, the traveling path of the road section on the two-dimensional map is determined, the future traveling track of the vehicle can be judged according to the current position of the vehicle through the road section information on the two-dimensional map, and the vehicle road is predicted in a self-adaptive mode.
4) Predicting future driving track and alarming
Predicting the future driving track of the vehicle on the two-dimensional map according to the vehicle driving historical track points and the vehicle yaw velocity of the vehicle, or judging the future driving track of the vehicle according to the ID number of the road section where the current vehicle is located and the current vehicle position on the two-dimensional map; the method for judging the future driving track of the vehicle is not limited; for example, the route of the road section in a future period of time, that is, the route on which the vehicle is going to travel, can be determined according to the current road section ID; or predicting the future running track of the vehicle according to the current vehicle position, the traveling direction and the vehicle yaw angular speed is a well-known vehicle path prejudging mode in the field of intelligent vehicles;
obtaining future driving tracks of surrounding object vehicles through vehicle communication;
when the object vehicle on the two-dimensional map is found to drive into the cross road section and has collision risk, judging whether the vehicle intersects with the path of the object vehicle in the three-dimensional space according to the ID number of the road section where the current vehicle is located and the ID number of the road section where the object vehicle is located;
intersecting and alarming; and the alarm is not given when the alarm is not intersected.
As shown in fig. 3, the lane prediction method for a vehicle, includes the steps of,
1) calling a lane-level map in vehicle information, wherein the lane-level map comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space layers of each road on the two-dimensional map;
2) collecting and storing historical tracks:
acquiring position and course information data of a group of vehicles in a period of time before the current time point through satellite positioning to form a group of historical track points of vehicle running;
3) judging the ID number of the road section where the current vehicle is located:
projecting a group of vehicle driving historical track points in the step 2) onto a two-dimensional map in the lane level map, wherein one vehicle driving historical track point corresponds to a plurality of road section ID numbers, and obtaining road section IDs with high frequency of occurrence of the road section ID numbers corresponding to the group of vehicle driving historical track points;
matching a track shape formed by historical track points of vehicle driving with road shapes of a plurality of road IDs in a two-dimensional map corresponding to the current vehicle position to obtain road shape matching road section ID numbers;
and weighting and fusing the ID numbers of the road shape matching road sections and the ID numbers of the road sections with high occurrence frequency, and selecting the ID numbers of the corresponding road sections with high matching degree and high occurrence frequency as the ID numbers of the road sections where the vehicles are located.
Therefore, the lane prediction is completed and can be used as a basis for judging the driving of the intelligent automobile, for example, the following step 4) is added for judging the collision.
4) Predicting future driving track and alarming
Predicting the future driving track of the vehicle on the two-dimensional map according to the vehicle driving historical track points and the vehicle yaw velocity of the vehicle, or judging the future driving track of the vehicle according to the ID number of the road section where the current vehicle is located and the current vehicle position on the two-dimensional map;
obtaining future driving tracks of surrounding object vehicles through vehicle communication;
when the object vehicle on the two-dimensional map is found to drive into the cross road section and has collision risk, judging whether the vehicle intersects with the path of the object vehicle in the three-dimensional space according to the ID number of the road section where the current vehicle is located and the ID number of the road section where the object vehicle is located;
intersecting and alarming; and the alarm is not given when the alarm is not intersected.
As shown in fig. 4, the lane prediction method for a vehicle, includes the steps of,
1) calling a lane-level map in vehicle information, wherein the lane-level map comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space layers of each road on the two-dimensional map;
2) collecting and storing historical tracks:
acquiring position and course information data of a group of vehicles in a period of time before the current time point through satellite positioning to form a group of historical track points of vehicle running;
3) judging the ID number of the road section where the current vehicle is located:
projecting a group of vehicle driving historical track points in the step 2) onto a two-dimensional map in the lane level map, wherein one vehicle driving historical track point corresponds to a plurality of road section ID numbers, and obtaining road section IDs with high frequency of occurrence of the road section ID numbers corresponding to the group of vehicle driving historical track points;
matching the current course of the vehicle with the road directions of a plurality of road IDs in a two-dimensional map corresponding to the current vehicle position to obtain road direction matching ID numbers;
and (4) weighting and fusing the road direction matching ID number and the road section ID number with high occurrence frequency, and selecting the corresponding road section ID number with high matching degree and high occurrence frequency as the road section ID number where the vehicle is located.
Therefore, the lane prediction is completed and can be used as a basis for judging the driving of the intelligent automobile, for example, the following step 4) is added for judging the collision.
4) Predicting future driving track and alarming
Predicting the future driving track of the vehicle on the two-dimensional map according to the vehicle driving historical track points and the vehicle yaw velocity of the vehicle, or judging the future driving track of the vehicle according to the ID number of the road section where the current vehicle is located and the current vehicle position on the two-dimensional map;
obtaining future driving tracks of surrounding object vehicles through vehicle communication;
when the object vehicle on the two-dimensional map is found to drive into the cross road section and has collision risk, judging whether the vehicle intersects with the path of the object vehicle in the three-dimensional space according to the ID number of the road section where the current vehicle is located and the ID number of the road section where the object vehicle is located;
intersecting and alarming; and the alarm is not given when the alarm is not intersected.
As shown in fig. 5, the lane prediction method for a vehicle, includes the steps of,
1) calling a lane-level map in vehicle information, wherein the lane-level map comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space layers of each road on the two-dimensional map;
2) collecting and storing historical tracks:
acquiring position and course information data of a group of vehicles in a period of time before the current time point through satellite positioning to form a group of historical track points of vehicle running;
3) judging the ID number of the road section where the current vehicle is located:
projecting a group of vehicle driving historical track points in the step 2) onto a two-dimensional map in the lane level map, wherein one vehicle driving historical track point corresponds to a plurality of road section ID numbers, and obtaining road section IDs with high frequency of occurrence of the road section ID numbers corresponding to the group of vehicle driving historical track points;
matching a track shape formed by historical track points of vehicle driving with road shapes of a plurality of road IDs in a two-dimensional map corresponding to the current vehicle position to obtain road shape matching road section ID numbers;
matching the current course of the vehicle with the road directions of a plurality of road IDs in a two-dimensional map corresponding to the current vehicle position to obtain road direction matching ID numbers;
weighting and fusing the road section ID numbers with high occurrence frequency, the road shape matching road section ID numbers and the road direction matching ID numbers, and selecting the corresponding road section ID numbers with high shape and direction matching degree and high occurrence frequency as the road section ID numbers where the vehicles are located;
therefore, the lane prediction is completed and can be used as a basis for judging the driving of the intelligent automobile, for example, the following step 4) is added for judging the collision.
4) Predicting future driving track and alarming
Predicting the future driving track of the vehicle on the two-dimensional map according to the vehicle driving historical track points and the vehicle yaw velocity of the vehicle, or judging the future driving track of the vehicle according to the ID number of the road section where the current vehicle is located and the current vehicle position on the two-dimensional map;
obtaining future driving tracks of surrounding object vehicles through vehicle communication;
when the object vehicle on the two-dimensional map is found to drive into the cross road section and has collision risk, judging whether the vehicle intersects with the path of the object vehicle in the three-dimensional space according to the ID number of the road section where the current vehicle is located and the ID number of the road section where the object vehicle is located;
intersecting and alarming; and the alarm is not given when the alarm is not intersected.
As shown in fig. 1, the lane-level map library module 3 stores a three-dimensional map library including a two-dimensional map and a unique link ID number capable of distinguishing a three-dimensional space hierarchy of each road on the two-dimensional map. The lane-level map in the lane-level map library module 3 is saved on the vehicle or obtained from a cloud server.
As shown in fig. 1, an automotive system includes: the system comprises a short-range wireless communication module 1, a satellite positioning module 2, a lane-level map library module 3 and a control unit module 4;
the short-range wireless communication module 1, the satellite positioning module 2 and the lane level map library module 3 are used as data input units and are in communication connection with the control unit module 4;
the short-range wireless communication module 1 receives the driving information (including position, course and speed information) sent by 7 surrounding object vehicles and sends the information to the control unit module 4;
the satellite positioning module 2 receives the driving information (including position, course and speed information) of the vehicle 6 by a satellite positioning method and sends the information to the control unit 4;
the lane-level map library module 3 is a three-dimensional map library, and each block comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space levels of each road on the two-dimensional map;
and the control unit module 4 is used for mapping the vehicle to a two-dimensional map according to the driving information of the satellite positioning module 2 to match a corresponding path and a road section ID number thereof, and judging the future driving track of the vehicle. The control unit module 4 may determine the collision between the host vehicle and the target vehicle based on the target vehicle travel information transmitted from the short range wireless communication module 1.
As shown in fig. 6, in a three-dimensional disjoint viaduct scene, the vehicle 6 runs on a road segment 1 under the viaduct, the target vehicle 7 runs on a road segment 2 on the viaduct, and according to the intersection collision early warning algorithm in the two-dimensional map of the existing intelligent vehicle technology, no consideration of spatial hierarchy exists, and the road segment 1 and the road segment 2 project to a two-dimensional map plane to have an intersection point, so that false alarm may occur; by adopting the technical scheme disclosed by the invention, the hierarchy of the current running road section of the vehicle (the hierarchy is formed at the road space hierarchy conversion part such as an overpass) can be judged by utilizing the historical track data and the road level map data, and the vehicle and the target vehicle are judged not to be in the same road section hierarchy without collision danger, so that false alarm is inhibited.
Claims (8)
1. The automobile lane prediction method is characterized by comprising the following steps: comprises the following steps of (a) carrying out,
1) calling a lane-level map in the vehicle information, wherein the lane-level map comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space layers of each road on the two-dimensional map;
2) collecting and storing historical tracks:
acquiring a group of vehicle position and course information data in a period of time before the current time point through satellite positioning to form a group of vehicle driving historical track points;
3) judging the ID number of the road section where the current vehicle is located:
projecting a group of vehicle driving historical track points in the step 2) onto a two-dimensional map in the lane level map, wherein one vehicle driving historical track point corresponds to a plurality of road section ID numbers, obtaining road section IDs with high frequency of occurrence of the road section ID numbers corresponding to the group of vehicle driving historical track points, and setting the road section ID with the highest frequency of occurrence as the road section ID number in the three-dimensional space where the current vehicle is located.
2. The automobile lane prediction method of claim 1, wherein: the method also comprises the following steps of,
4) predicting future driving track and alarming
Predicting the future driving track of the vehicle on the two-dimensional map according to the vehicle driving historical track points and the vehicle yaw velocity of the vehicle, or judging the future driving track of the vehicle according to the ID number of the road section where the current vehicle is located and the current vehicle position on the two-dimensional map;
obtaining future driving tracks of surrounding object vehicles through vehicle communication;
when the object vehicle on the two-dimensional map is found to drive into the cross road section and has collision risk, judging whether the vehicle intersects with the path of the object vehicle in the three-dimensional space according to the ID number of the road section where the current vehicle is located and the ID number of the road section where the object vehicle is located;
intersecting and alarming; and the alarm is not given when the alarm is not intersected.
3. The automobile lane prediction method is characterized by comprising the following steps: comprises the following steps of (a) carrying out,
1) calling a lane-level map in the vehicle information, wherein the lane-level map comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space layers of each road on the two-dimensional map;
2) collecting and storing historical tracks:
acquiring position and course information data of a group of vehicles in a period of time before the current time point through satellite positioning to form a group of historical track points of vehicle running;
3) judging the ID number of the road section where the current vehicle is located:
projecting a group of vehicle driving historical track points in the step 2) onto a two-dimensional map in the lane level map, wherein one vehicle driving historical track point corresponds to a plurality of road section ID numbers, and obtaining road section IDs with high frequency of occurrence of the road section ID numbers corresponding to the group of vehicle driving historical track points;
matching a track shape formed by historical track points of vehicle driving with road shapes of a plurality of road IDs in a two-dimensional map corresponding to the current vehicle position to obtain road shape matching road section ID numbers;
fusing and judging the ID number of the road section with the shape matching road section and the ID number of the road section with high occurrence frequency to obtain the ID number of the road section where the current vehicle is located; and the fusion judgment is a weighted fusion method, and the ID number of the corresponding road section with high matching degree and high occurrence frequency is selected as the ID number of the road section where the vehicle is located.
4. The automobile lane prediction method of claim 3, wherein: the method also comprises the following steps of,
4) predicting future driving track and alarming
Predicting the future driving track of the vehicle on the two-dimensional map according to the vehicle driving historical track points and the vehicle yaw velocity of the vehicle, or judging the future driving track of the vehicle according to the ID number of the road section where the current vehicle is located and the current vehicle position on the two-dimensional map;
obtaining future driving tracks of surrounding object vehicles through vehicle communication;
when the object vehicle on the two-dimensional map is found to drive into the cross road section and has collision risk, judging whether the vehicle intersects with the path of the object vehicle in the three-dimensional space according to the ID number of the road section where the current vehicle is located and the ID number of the road section where the object vehicle is located;
intersecting and alarming; and the alarm is not given when the alarm is not intersected.
5. The automobile lane prediction method is characterized by comprising the following steps: comprises the following steps of (a) carrying out,
1) calling a lane-level map in the vehicle information, wherein the lane-level map comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space layers of each road on the two-dimensional map;
2) collecting and storing historical tracks:
acquiring position and course information data of a group of vehicles in a period of time before the current time point through satellite positioning to form a group of historical track points of vehicle running;
3) judging the ID number of the road section where the current vehicle is located:
projecting a group of vehicle driving historical track points in the step 2) onto a two-dimensional map in the lane level map, wherein one vehicle driving historical track point corresponds to a plurality of road section ID numbers, and obtaining road section IDs with high frequency of occurrence of the road section ID numbers corresponding to the group of vehicle driving historical track points;
matching the current course of the vehicle with the road directions of a plurality of road IDs in a two-dimensional map corresponding to the current vehicle position to obtain road direction matching ID numbers;
merging and judging the road direction matching ID number and the road section ID number with high occurrence frequency to obtain the road section ID number of the current vehicle; and the fusion judgment is a weighted fusion method, and the ID number of the corresponding road section with high matching degree and high occurrence frequency is selected as the ID number of the road section where the vehicle is located.
6. The automobile lane prediction method of claim 5, wherein: the method also comprises the following steps of,
4) predicting future driving track and alarming
Predicting the future driving track of the vehicle on the two-dimensional map according to the vehicle driving historical track points and the vehicle yaw velocity of the vehicle, or judging the future driving track of the vehicle according to the ID number of the road section where the current vehicle is located and the current vehicle position on the two-dimensional map;
obtaining future driving tracks of surrounding object vehicles through vehicle communication;
when the object vehicle on the two-dimensional map is found to drive into the cross road section and has collision risk, judging whether the vehicle intersects with the path of the object vehicle in the three-dimensional space according to the ID number of the road section where the current vehicle is located and the ID number of the road section where the object vehicle is located;
intersecting and alarming; and the alarm is not given when the alarm is not intersected.
7. The automobile lane prediction method is characterized by comprising the following steps: comprises the following steps of (a) carrying out,
1) calling a lane-level map in the vehicle information, wherein the lane-level map comprises a two-dimensional map and a unique road section ID number capable of distinguishing three-dimensional space layers of each road on the two-dimensional map;
2) collecting and storing historical tracks:
acquiring position and course information data of a group of vehicles in a period of time before the current time point through satellite positioning to form a group of historical track points of vehicle running;
3) judging the ID number of the road section where the current vehicle is located:
projecting a group of vehicle driving historical track points in the step 2) onto a two-dimensional map in the lane level map, wherein one vehicle driving historical track point corresponds to a plurality of road section ID numbers, and obtaining road section IDs with high frequency of occurrence of the road section ID numbers corresponding to the group of vehicle driving historical track points;
matching a track shape formed by historical track points of vehicle driving with road shapes of a plurality of road IDs in a two-dimensional map corresponding to the current vehicle position to obtain road shape matching road section ID numbers;
matching the current course of the vehicle with the road directions of a plurality of road IDs in a two-dimensional map corresponding to the current vehicle position to obtain road direction matching ID numbers;
fusing and judging the ID number of the road section with high frequency, the ID number of the road shape matching road section and the ID number of the road direction matching, and obtaining the ID number of the road section where the current vehicle is located; and the fusion judgment is a weighted fusion method, and the ID number of the corresponding road section with high shape and direction matching degree and high occurrence frequency is selected as the ID number of the road section where the vehicle is located.
8. The automobile lane prediction method of claim 7, wherein: the method also comprises the following steps of,
4) predicting future driving track and alarming
Predicting the future driving track of the vehicle on the two-dimensional map according to the vehicle driving historical track points and the vehicle yaw velocity of the vehicle, or judging the future driving track of the vehicle according to the ID number of the road section where the current vehicle is located and the current vehicle position on the two-dimensional map;
obtaining future driving tracks of surrounding object vehicles through vehicle communication;
when the object vehicle on the two-dimensional map is found to drive into the cross road section and has collision risk, judging whether the vehicle intersects with the path of the object vehicle in the three-dimensional space according to the ID number of the road section where the current vehicle is located and the ID number of the road section where the object vehicle is located;
intersecting and alarming; and the alarm is not given when the alarm is not intersected.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710384424.6A CN107221195B (en) | 2017-05-26 | 2017-05-26 | Automobile lane prediction method and lane level map |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710384424.6A CN107221195B (en) | 2017-05-26 | 2017-05-26 | Automobile lane prediction method and lane level map |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107221195A CN107221195A (en) | 2017-09-29 |
| CN107221195B true CN107221195B (en) | 2020-03-17 |
Family
ID=59945458
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710384424.6A Active CN107221195B (en) | 2017-05-26 | 2017-05-26 | Automobile lane prediction method and lane level map |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107221195B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108629978B (en) * | 2018-06-07 | 2020-12-22 | 重庆邮电大学 | Traffic track prediction method based on high-dimensional road network and recurrent neural network |
| EP3640679B1 (en) | 2018-10-15 | 2023-06-07 | Zenuity AB | A method for assigning ego vehicle to a lane |
| CN110174892B (en) * | 2019-04-08 | 2022-07-22 | 阿波罗智能技术(北京)有限公司 | Vehicle orientation processing method, device, equipment and computer readable storage medium |
| US20220172618A1 (en) * | 2019-04-12 | 2022-06-02 | Continental Automotive Systems, Inc. | Electronic Control Device For A Vehicle And Method For Reducing Intersection False-Positive Detection |
| CN110489510B (en) * | 2019-08-23 | 2022-05-20 | 腾讯科技（深圳）有限公司 | Road data processing method and device, readable storage medium and computer equipment |
| CN111337045A (en) * | 2020-03-27 | 2020-06-26 | 北京百度网讯科技有限公司 | Vehicle navigation method and device |
| CN111626097A (en) * | 2020-04-09 | 2020-09-04 | 吉利汽车研究院（宁波）有限公司 | Method and device for predicting future trajectory of obstacle, electronic equipment and storage medium |
| CN112017428B (en) * | 2020-07-09 | 2021-12-17 | 惠州市德赛西威智能交通技术研究院有限公司 | Road side vehicle networking device, viaduct road section identification method and vehicle-mounted vehicle networking device |
| CN112130137B (en) * | 2020-09-17 | 2023-10-20 | 杭州海康威视数字技术股份有限公司 | Method, device and storage medium for determining lane-level track |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101788298A (en) * | 2008-10-31 | 2010-07-28 | 歌乐牌株式会社 | Guider and air navigation aid |
| CN103175536A (en) * | 2011-12-22 | 2013-06-26 | 罗伯特·博世有限公司 | Method for displaying object on display of navigation system in simply and 3-D manner |
| CN103557870A (en) * | 2013-10-09 | 2014-02-05 | 董路 | Dynamic trajectory navigation method and cloud platform |
| CN104680838A (en) * | 2013-11-28 | 2015-06-03 | 三星电子（中国）研发中心 | Safety assisting method and system for automobile |
| CN104916133A (en) * | 2015-06-09 | 2015-09-16 | 福建工程学院 | Road altitude information extraction method and system based on traffic track data |
| CN105893565A (en) * | 2016-03-31 | 2016-08-24 | 百度在线网络技术（北京）有限公司 | Method and device for establishing a three-dimensional map database and transmitting three-dimensional map data |
| CN106251699A (en) * | 2016-08-19 | 2016-12-21 | 深圳市元征科技股份有限公司 | Vehicle running collision method for early warning and device |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2959684A4 (en) * | 2013-02-20 | 2016-11-09 | Intel Corp | Real-time automatic conversion of 2-dimensional images or video to 3-dimensional stereo images or video |
| US20160092068A1 (en) * | 2014-09-30 | 2016-03-31 | International Business Machines Corporation | Visualization of addresses |
-
2017
- 2017-05-26 CN CN201710384424.6A patent/CN107221195B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101788298A (en) * | 2008-10-31 | 2010-07-28 | 歌乐牌株式会社 | Guider and air navigation aid |
| CN103175536A (en) * | 2011-12-22 | 2013-06-26 | 罗伯特·博世有限公司 | Method for displaying object on display of navigation system in simply and 3-D manner |
| CN103557870A (en) * | 2013-10-09 | 2014-02-05 | 董路 | Dynamic trajectory navigation method and cloud platform |
| CN104680838A (en) * | 2013-11-28 | 2015-06-03 | 三星电子（中国）研发中心 | Safety assisting method and system for automobile |
| CN104916133A (en) * | 2015-06-09 | 2015-09-16 | 福建工程学院 | Road altitude information extraction method and system based on traffic track data |
| CN105893565A (en) * | 2016-03-31 | 2016-08-24 | 百度在线网络技术（北京）有限公司 | Method and device for establishing a three-dimensional map database and transmitting three-dimensional map data |
| CN106251699A (en) * | 2016-08-19 | 2016-12-21 | 深圳市元征科技股份有限公司 | Vehicle running collision method for early warning and device |
Non-Patent Citations (1)
| Title |
|---|
| 动态车辆导航系统车道级路径引导方法研究;张林;《中国博士学位论文全文数据库 工程科技Ⅱ辑》;20081115;正文第27-53页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107221195A (en) | 2017-09-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107221195B (en) | Automobile lane prediction method and lane level map | |
| US10963462B2 (en) | Enhancing autonomous vehicle perception with off-vehicle collected data | |
| US8160811B2 (en) | Method and system to estimate driving risk based on a hierarchical index of driving | |
| CN102473346B (en) | Vehicle controller, control method for vehicle and control system for vehicle | |
| JP6462328B2 (en) | Travel control system | |
| CN107792077B (en) | Method and system for confirming whether road section is suitable for autonomous vehicle driving | |
| Lytrivis et al. | An advanced cooperative path prediction algorithm for safety applications in vehicular networks | |
| US20120303222A1 (en) | Driver assistance system | |
| WO2018233175A1 (en) | Danger early-warning method and system, v2x vehicle-mounted terminal and memory | |
| US20120245756A1 (en) | Driver assistance system | |
| Kim et al. | Design of integrated risk management-based dynamic driving control of automated vehicles | |
| JP7328759B2 (en) | Contextual transmission of MAP messages to improve digital maps | |
| CN110562222B (en) | Emergency braking control method for curve scene, vehicle-mounted device and storage medium | |
| JP2016192150A (en) | Vehicle travel control device | |
| CN110807412B (en) | Vehicle laser positioning method, vehicle-mounted equipment and storage medium | |
| US11731653B2 (en) | Conditional motion predictions | |
| CN106355890A (en) | Method and device for judging classification of target vehicle | |
| US11042160B2 (en) | Autonomous driving trajectory determination device | |
| CN107200016B (en) | Road adaptive forecasting method and the Vehicular system for using this method | |
| CN111051817A (en) | Method for determining the position of a motor vehicle in the environment, control device for a motor vehicle and computing device for operating on a data network | |
| CN109515436B (en) | Selection method for intelligent driving vehicle running path fusion | |
| CN113873426A (en) | System, control unit and method for deciding on a geo-fence event of a vehicle | |
| CN114126940A (en) | Electronic control device | |
| CN109815555A (en) | The environmental modeling capability assessment method and system of automatic driving vehicle | |
| JP2020125108A (en) | Lane detection method and system for vehicle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20170929 Assignee: CHONGQING CHANGAN NEW ENERGY AUTOMOBILE TECHNOLOGY Co.,Ltd. Assignor: Chongqing Changan Automobile Co.,Ltd. Contract record no.: X2021500000014 Denomination of invention: Vehicle lane prediction method and lane level map Granted publication date: 20200317 License type: Common License Record date: 20211014 |