Search

CN-122009168-A - Lane selection method, lane selection device, vehicle and readable storage medium

CN122009168ACN 122009168 ACN122009168 ACN 122009168ACN-122009168-A

Abstract

The embodiment of the application provides a lane selection method, a lane selection device, a vehicle and a readable storage medium. The method comprises the steps of obtaining road information of a position of a vehicle, wherein the road information comprises lane information of a current lane, an ultra-wide road section and a subsequent road section of the vehicle, determining a plurality of lane sequences based on the road information, wherein one lane sequence comprises the current lane, one virtual lane and one candidate lane, selecting a target virtual lane from the plurality of virtual lanes based on first lane selection costs of each lane sequence if the vehicle is about to drive into the ultra-wide road section, controlling the vehicle to drive into the target virtual lane, determining second lane selection costs of each candidate lane if the vehicle is about to drive into the subsequent road section from the target virtual lane, selecting the target candidate lane from the plurality of candidate lanes based on the second lane selection costs of each candidate lane, and controlling the vehicle to drive into the target candidate lane. The application can select a more reasonable lane from the road section with unclear lane lines or completely missing lane lines.

Inventors

  • LI XIUYUAN
  • TU NINGNING
  • HOU MINGYANG
  • YIN LEI
  • LIU WAN

Assignees

  • 岚图汽车科技股份有限公司

Dates

Publication Date
20260512
Application Date
20260108

Claims (17)

  1. 1. A lane selection method, the method comprising: The method comprises the steps that road information of a position where a vehicle is located is obtained, the road information comprises a current lane where the vehicle is located, an ultra-wide road section connected with the current lane and lane information of a subsequent road section connected with the ultra-wide road section, the ultra-wide road section comprises a plurality of virtual lanes, and the subsequent road section comprises a plurality of candidate lanes; Determining a plurality of lane sequences based on the road information, one lane sequence including the current lane, one virtual lane, and one candidate lane; If the vehicle is about to drive into the ultra-wide road section, selecting a target virtual lane from the plurality of virtual lanes based on first lane selection cost of each lane sequence, and controlling the vehicle to drive into the target virtual lane; If the vehicle is about to drive into the subsequent road section from the target virtual lane, determining second lane selection costs of all candidate lanes, wherein the same pre-aiming point is adopted when determining the first lane selection cost of the lane sequence and the second lane selection cost of the corresponding candidate lane, and the pre-aiming point is positioned in the candidate lane in the lane sequence; Selecting a target lane candidate from the plurality of lane candidates based on the second lane selection cost of each lane candidate, and controlling the vehicle to drive into the target lane candidate.
  2. 2. The method of claim 1, wherein the first lane cost is determined based on a first lateral projection cost, a first split trend cost, and a first steering consistency cost of the lane sequence; The second lane selection cost is determined based on a second lateral projection cost, a second diversion trend cost, and a second diversion consistency cost of the candidate lane; when the first steering consistency cost of the lane sequence and the second steering consistency cost of the corresponding candidate lane are determined, the same pre-aiming point is adopted, and the pre-aiming point is positioned in the candidate lane in the lane sequence.
  3. 3. The method of claim 2, wherein the lane information of the ultra-wide road segment includes lane information of each of the virtual lanes, the method further comprising: for each lane sequence, determining the first transverse projection cost according to the lane information of the current lane and the lane information of the virtual lane in the lane sequence; determining the first diversion trend cost according to the lane information of the current lane and the lane information of the virtual lane in the lane sequence; Determining the first steering consistency cost according to the actual steering wheel angle of the vehicle and the pretightening point corresponding to the lane sequence; And determining the first lane selection cost of the lane sequence according to the first transverse projection cost, the first diversion trend cost and the first steering consistency cost.
  4. 4. The method of claim 3, wherein the lane information for the current lane comprises centerline information for the current lane and the lane information for the virtual lane comprises centerline information for the virtual lane, wherein the determining the first lateral projection cost based on the lane information for the current lane and the lane information for the virtual lane in the lane sequence comprises: determining the transverse projection distance between the current lane and the virtual lane according to the central line information of the current lane and the central line information of the virtual lane; and determining the first transverse projection cost according to the transverse projection distance between the current lane and the virtual lane.
  5. 5. The method of claim 4, wherein the determining the lateral projected distance of the current lane from the virtual lane based on the centerline information of the current lane and the centerline information of the virtual lane comprises: Under the condition that the road section type of the ultra-wide road section is a parallel ultra-wide type, taking the transverse projection distance of a first position relative to a second position as the transverse projection distance of the current lane and the virtual lane, wherein the first position is an end point of the central line of the current lane, and the second position is a starting point of the central line of the virtual lane; The determining the first lateral projection cost according to the lateral projection distance between the current lane and the virtual lane includes: and taking the transverse projection distance between the current lane and the virtual lane as the first transverse projection cost.
  6. 6. The method of claim 4, wherein the lateral projected distance of the current lane from the virtual lane comprises a first lateral projected distance and a second lateral projected distance, wherein the determining the lateral projected distance of the current lane from the centerline information of the current lane and the centerline information of the virtual lane comprises: Taking a transverse projection distance of a first position relative to a second position as the first transverse projection distance under the condition that the road section type of the ultra-wide road section is a Y-type ultra-wide type, wherein the first position is an end point of the central line of the current lane, and the second position is a starting point of the central line of the virtual lane; Taking a transverse projection distance of a third position relative to the central line of the virtual lane as the second transverse projection distance, wherein the third position is positioned in the extending direction of the central line of the current lane, and the distance between the third position and the first position is a first preset distance; The determining the first lateral projection cost according to the lateral projection distance between the current lane and the virtual lane includes: And determining the first transverse projection cost according to the first transverse projection distance and the second transverse projection distance.
  7. 7. The method of claim 3, wherein the lane information for the current lane comprises direction information for the current lane and the lane information for the virtual lane comprises direction information for the virtual lane, wherein the determining the first split trend cost based on the lane information for the current lane and the lane information for the virtual lane in the lane sequence comprises: determining a first angle between the current lane and the virtual lane according to the direction information of the current lane and the direction information of the virtual lane; and determining the first diversion trend cost according to the first angle, wherein the first diversion trend cost is positively correlated with the absolute value of the first angle.
  8. 8. The method of claim 7, wherein said determining said first split trend cost from said first angle comprises: if the absolute value of the first angle is smaller than a preset angle threshold, determining the first diversion trend cost according to the first angle; And if the absolute value of the first angle is greater than or equal to the preset angle threshold, determining the first diversion trend cost according to the preset angle threshold.
  9. 9. A method according to claim 3, wherein said determining said first steering consistency cost from an actual steering wheel angle of said vehicle and a pretightening point corresponding to said lane sequence comprises: determining a first reference front wheel corner according to a pre-aiming point corresponding to the lane sequence; determining a first reference steering wheel angle according to the first reference front wheel angle and a transmission ratio of the vehicle; and determining the first steering consistency cost according to the actual steering wheel angle and the first reference steering wheel angle of the vehicle.
  10. 10. The method of claim 3, wherein the determining the first lane cost of the lane sequence based on the first lateral projection cost, the first split trend cost, and the first turn consistency cost comprises: and carrying out weighted average processing on the first transverse projection cost, the first diversion trend cost and the first steering consistency cost, and determining the first lane selection cost of the lane sequence.
  11. 11. The method of claim 2, wherein the lane information of the subsequent road segment includes lane information of each of the candidate lanes, and wherein the determining the second lane-selection cost of each of the candidate lanes comprises: Determining the second transverse projection cost according to the lane information of the target virtual lane and the lane information of the candidate lane for each candidate lane; determining the second shunt trend cost according to the lane information of the target virtual lane and the lane information of the candidate lane; Determining the second steering consistency cost according to the actual steering wheel angle of the vehicle and a pretightening point corresponding to a lane sequence where the candidate lane is located; And determining the second lane selection cost of the candidate lane according to the second transverse projection cost, the second diversion trend cost and the second steering consistency cost.
  12. 12. The method according to claim 1, wherein the method further comprises: and determining pretightening points corresponding to each lane sequence based on the lane information of the subsequent road section.
  13. 13. The method of claim 12, wherein the determining the pretightening point corresponding to each lane sequence based on the lane information of the subsequent road segment comprises: If a preset type of lanes exist in a plurality of candidate lanes of the subsequent road section, taking a starting point of a central line of the candidate lanes as a preset aiming point of a lane sequence where the candidate lanes are located, wherein the preset type of lanes are not ultra-wide lanes.
  14. 14. The method of claim 13, wherein the subsequent road segment comprises a plurality of sub-road segments, each sub-road segment comprising a plurality of candidate lanes, the determining the pre-aiming point for each lane sequence based on lane information for the subsequent road segment comprising: If a first sub-road section exists in the plurality of sub-road sections, taking a starting point of a central line of a candidate lane in the first sub-road section as a pre-aiming point of a lane sequence where the candidate lane is located, wherein the first sub-road section is a sub-road section with the preset type lane in the plurality of sub-road sections; and if the first sub-road section does not exist in the plurality of sub-road sections, taking the starting point of the central line of the candidate lane in a second sub-road section as a pre-aiming point of the lane sequence where the candidate lane is located, wherein the second sub-road section is the last sub-road section in the plurality of sub-road sections.
  15. 15. A lane selection apparatus, comprising: The system comprises an acquisition module, a control module and a control module, wherein the acquisition module is used for acquiring road information of a position of a vehicle, the road information comprises a current lane of the vehicle, an ultra-wide road section connected with the current lane and lane information of a subsequent road section connected with the ultra-wide road section, the ultra-wide road section comprises a plurality of virtual lanes, and the subsequent road section comprises a plurality of candidate lanes; A first determining module configured to determine a plurality of lane sequences, one lane sequence including the current lane, one virtual lane, and one candidate lane, based on the road information; The first lane selection module is used for selecting a target virtual lane from the plurality of virtual lanes based on first lane selection cost of each lane sequence and controlling the vehicle to drive into the target virtual lane if the vehicle is about to drive into the ultra-wide road section; A second determining module, configured to determine a second lane selection cost of each candidate lane if the vehicle is about to drive into the subsequent road segment from the target virtual lane, where the same pre-aiming point is adopted when determining the first lane selection cost of the lane sequence and the second lane selection cost of the corresponding candidate lane, where the pre-aiming point is located in the candidate lane in the lane sequence; and the second lane selection module is used for selecting a target candidate lane from the plurality of candidate lanes based on the second lane selection cost of each candidate lane and controlling the vehicle to drive into the target candidate lane.
  16. 16. A vehicle is characterized by comprising a memory and a processor; The memory stores computer-executable instructions; the processor executing computer-executable instructions stored in the memory, causing the processor to perform the method of any one of claims 1-14.
  17. 17. A computer readable storage medium having stored therein computer executable instructions which when executed by a processor are adapted to carry out the method of any one of claims 1-14.

Description

Lane selection method, lane selection device, vehicle and readable storage medium Technical Field The present application relates to the field of assisted driving technologies, and in particular, to a lane selection method, a lane selection device, a vehicle, and a readable storage medium. Background Lane centering (LANE CENTERING control, LCC) is an advanced driving assistance system function that senses the lane line of the road where the vehicle is located in real time through a sensor (e.g., an on-vehicle camera, a laser radar, etc.), and adjusts the steering wheel based on the sensing result, so that the vehicle automatically runs along the current lane. In an actual driving scenario, when a vehicle passes through an intersection (such as an intersection, a split intersection, or a widened intersection), a lane selection is required to guide the vehicle into the next driving path with the selected lane as a tracking target. However, in the related art, the lane selection strategy is severely dependent on the perceived lane line. For a road section with unclear lane lines or completely missing lane lines, accurate and stable virtual lane lines are difficult to generate, so that a more reasonable target lane is difficult to select, or the problem that the selected lane frequently jumps occurs, and the driving safety and the comfort of a vehicle are affected. Disclosure of Invention The embodiment of the application provides a lane selection method, a lane selection device, a vehicle and a readable storage medium, which can select a more reasonable target virtual lane from a road section with unclear or completely missing lane lines by moving a pre-aiming point backwards to a geometrically stable subsequent road section, can avoid the problem of frequent jump of the pre-aiming point caused by the geometrical instability of the virtual lane in an ultra-wide road section, and can ensure the stability of the selected lane. The method comprises the steps of obtaining road information of a position of a vehicle, wherein the road information comprises a current lane of the vehicle, an ultra-wide road section connected with the current lane and lane information of a subsequent road section connected with the ultra-wide road section, the ultra-wide road section comprises a plurality of virtual lanes, the subsequent road section comprises a plurality of candidate lanes, a plurality of lane sequences are determined based on the road information, one lane sequence comprises the current lane, one virtual lane and one candidate lane, a target virtual lane is selected from the plurality of virtual lanes based on first lane selecting cost of each lane sequence if the vehicle is about to drive into the ultra-wide road section, and the vehicle is controlled to drive into the target virtual lane, second lane selecting cost of each candidate lane is determined if the vehicle is about to drive into the subsequent road section from the target virtual lane, the same pre-targeting point is adopted when the first lane selecting cost of each lane sequence and the second lane selecting cost of the corresponding candidate lane are determined, the candidate lane is pre-targeting point is pre-targeting the candidate lanes in the lane sequence, and the target lane candidates are selected from the plurality of target lane candidates based on the second lane selecting cost of each candidate lane. In one possible implementation manner, the first lane selection cost is determined based on a first transverse projection cost, a first diversion trend cost and a first steering consistency cost of the lane sequence, and the second lane selection cost is determined based on a second transverse projection cost, a second diversion trend cost and a second steering consistency cost of the candidate lane, wherein the same pre-aiming point is adopted when determining the first steering consistency cost of the lane sequence and the second steering consistency cost of the corresponding candidate lane, and the pre-aiming point is used for pre-aiming the candidate lane in the lane sequence. In a possible implementation manner, the lane information of the ultra-wide road section comprises lane information of each virtual lane, the method further comprises determining, for each lane sequence, a first transverse projection cost according to the lane information of the current lane and the lane information of the virtual lanes in the lane sequence, determining a first diversion trend cost according to the lane information of the current lane and the lane information of the virtual lanes in the lane sequence, determining a first steering consistency cost according to an actual steering wheel angle of the vehicle and a pre-aiming point corresponding to the lane sequence, and determining a first lane selection cost of the lane sequence according to the first transverse projection cost, the first diversion trend cost and the first steering consistency cost. In o