Search

EP-4502542-B1 - PATH PLANNING METHOD AND APPARATUS AND STORAGE MEDIUM

EP4502542B1EP 4502542 B1EP4502542 B1EP 4502542B1EP-4502542-B1

Inventors

  • LI, CHANGCHUN

Dates

Publication Date
20260506
Application Date
20230310

Claims (10)

  1. A path planning method, comprising: projecting (S301) information on lane lines, information on a lane center-line and information on an obstacle contour into a target coordinate system to obtain projection information; determining (S302) a start point and an end point of path planning in the target coordinate system, and if it is determined based on the projection information that there is an obstacle between the start point and the end point, then performing sampling on both sides of the obstacle contour along a travelling direction of a vehicle to obtain sampling points; determining (S303) a plurality of paths based on the start point, the end point and the sampling points; performing (S304) a cost calculation on the plurality of paths, and determining a reference path in the plurality of paths based on results of the cost calculation, wherein the results of the cost calculation characterize degrees to which the paths are close to the lane center-line in a case of avoiding the obstacle; and determining (S305) a target planned path based on the reference path; wherein performing (S304) the cost calculation on the plurality of paths, and determining the reference path in the plurality of paths based on the results of the cost calculation comprise: for each path of the plurality of paths, determining a first distance between a point on the path and the lane center-line, and a second distance between the point on the path and the obstacle contour, and determining a result of the cost calculation corresponding to the path according to the first distance, the second distance and a smooth-property value of the path; and comparing the results of the cost calculation corresponding to the paths, and determining a path of which the result of the cost calculation is minimal as the reference path; wherein determining the result of the cost calculation corresponding to the path according to the first distance, the second distance and the smooth-property value of the path comprises: for each point on the path, obtaining a first calculation result based on a product of the first distance corresponding to the point and a first coefficient, and obtaining a second calculation result based on a product of the second distance corresponding to the point and a second coefficient; performing a linear fusion based on the first coefficient and the second coefficient to obtain a third coefficient, and obtaining a third calculation result based on a product of the smooth-property value of the path and the third coefficient; summing the first calculation result, the second calculation result and the third calculation result corresponding to each point on the path to obtain a first summing result; and adjusting the first coefficient and the second coefficient based on the first summing result, and in a case that the second distance between each point on the path and the obstacle contour is not less than a safety distance, determining the result of the cost calculation corresponding to the path.
  2. The path planning method according to claim 1, wherein performing the sampling on both sides of the obstacle contour along the travelling direction of the vehicle to obtain the sampling points comprises: determining a sampling distance on both sides of the obstacle contour along the travelling direction of the vehicle, wherein the sampling distance is a sum of half a width of the vehicle and a safety distance, and the safety distance is a preset distance between the vehicle and the obstacle contour; and performing sampling based on the sampling distance to obtain the sampling points.
  3. The path planning method according to claim 2, wherein the sampling points comprise a valid sampling point and an invalid sampling point, wherein the valid sampling point is located within the lane lines, and the invalid sampling point is located outside the lane lines.
  4. The path planning method according to claim 3, further comprising: if all the sampling points are invalid sampling points, then taking a point of intersection between a lower boundary of the obstacle contour and the lane center-line as a final sampling point; and taking the final sampling point as the end point.
  5. The path planning method according to claim 1, further comprising: if it is determined based on the projection information that there is no obstacle between the start point and the end point, then determining the lane center-line as the reference path.
  6. A path planning apparatus, comprising: a projecting module (701), configured to project information on lane lines, information on a lane center-line and information on an obstacle contour into a target coordinate system to obtain projection information; a sampling module (702), configured to determine a start point and an end point of path planning in the target coordinate system, and if it is determined based on the projection information that there is an obstacle between the start point and the end point, then perform sampling on both sides of the obstacle contour along a travelling direction of a vehicle to obtain sampling points; a planning module (703), configured to determine a plurality of paths based on the start point, the end point and the sampling points; a cost calculating module (704), configured to perform a cost calculation on the plurality of paths, and determine a reference path in the plurality of paths based on results of the cost calculation, wherein the results of the cost calculation characterize degrees to which the paths are close to the lane center-line in a case of avoiding the obstacle; and an optimizing module (705), configured to determine a target planned path based on the reference path; wherein the cost calculating module (704) comprises: a parameter determining module, configured to: for each path of the plurality of paths, determine a first distance between a point on the path and the lane center-line, and a second distance between the point on the path and the obstacle contour; a calculating module, configured to: for each path of the plurality of paths, determine a result of the cost calculation corresponding to the path according to the first distance, the second distance and a smooth-property value of the path; and a reference-path determining module, configured to compare the results of the cost calculation corresponding to the paths, and determine a path of which the result of the cost calculation is minimal as the reference path; wherein the calculating module comprises: a first calculating unit, configured to: for each point on the path, obtain a first calculation result based on a product of the first distance corresponding to the point and a first coefficient, and obtain a second calculation result based on a product of the second distance corresponding to the point and a second coefficient; a second calculating unit, configured to perform a linear fusion based on the first coefficient and the second coefficient to obtain a third coefficient, and obtain a third calculation result based on a product of the smooth-property value of the path and the third coefficient; a third calculating unit, configured to sum the first calculation result, the second calculation result and the third calculation result corresponding to each point on the path to obtain a first summing result; and an adjusting unit, configured to adjust the first coefficient and the second coefficient based on the first summing result, and in a case that the second distance between each point on the path and the obstacle contour is not less than a safety distance, determine the result of the cost calculation corresponding to the path.
  7. The path planning apparatus according to claim 6, wherein the sampling module (702) comprises: a sampling-distance determining unit, configured to determine a sampling distance on both sides of the obstacle contour along the travelling direction of the vehicle, wherein the sampling distance is a sum of half a width of the vehicle and a safety distance, and the safety distance is a preset distance between the vehicle and the obstacle contour; and a sampling unit, configured to perform sampling based on the sampling distance to obtain the sampling points.
  8. The path planning apparatus according to claim 7, wherein the sampling points comprise a valid sampling point and an invalid sampling point, wherein the valid sampling point is located within the lane lines, and the invalid sampling point is located outside the lane lines.
  9. The path planning apparatus according to claim 8, wherein the sampling module (702) is configured to: if all the sampling points are invalid sampling points, then take a point of intersection between a lower boundary of the obstacle contour and the lane center-line as a final sampling point; and take the final sampling point as the end point.
  10. A computer-readable storage medium, wherein at least one instruction or at least one program is stored in the computer-readable storage medium, and the at least one instruction or the at least one program is loaded and executed by a processor to implement the steps of the method according to any one of claims 1 to 5.

Description

TECHNICAL FIELD The invention relates to the field of intelligent transportation, in particular to a path planning method, a path planning apparatus, an electronic device and a storage medium. BACKGROUND At present, there are three main types of path planning manners in mainstream: graph search, sampling, and numerical optimization, with different advantages and disadvantages. The graph search can ensure that a searched-out path is optimal, and the numerical optimization can provide a good guarantee of a smooth path. Referring to FIG. 1, a path planning method in the prior art is usually that: a self-driving vehicle firstly searches for a reference path having no collision with an obstacle in a hard boundary, where this path will be in an expected safe travelling corridor and close to a lane center-line of the current lane as much as possible by designing a cost function; a path meeting a vehicle dynamics constraint is then planned in the hard boundary based on the reference path by means of the numerical optimization; and finally the vehicle travels following the planned path. For the self-driving vehicle, the vehicle needs to plan a safe and collision-free path that meets a vehicle kinematic constraint. Referring to the left of FIG. 2, a current common practice to ensure a path collision-free with an obstacle is to solve such problem through obstacle inflation, and determination on a direction for avoiding the obstacle depends on a rule decision. Referring to the right of FIG. 2, inflation with a fixed size for the obstacle makes it impossible for the self-driving vehicle to plan a path when a travelling space is narrow, and the rule-based determination cannot effectively and correctly calculate directions for avoiding multiple obstacles. CN113124891A discloses a driving path planning method, which includes: obtaining a starting point S, an ending point E, and an obstacle set; the obstacle set contains all obstacles within the effective traffic area; connecting the starting point S and the ending point E to generate a vector SE; when the distance from any obstacle in the obstacle set to the vector SE is less than the traffic width, generating two path points on both sides of each obstacle in the obstacle set perpendicular to the direction of the vector SE; and obtaining the driving path according to the two path points corresponding to each obstacle in the obstacle set, the starting point S, and the ending point E in the direction of the vector SE. US2020385017A1 discloses a vehicle control device, which includes an obstacle recognizer configured to recognize an obstacle which is located near a vehicle and a target trajectory generator configured to generate a target trajectory in which the vehicle is to travel repeatedly with a predetermined cycle, where the target trajectory generator is configured to generate the target trajectory such that a first change which is an amount of change in a road width direction from the target trajectory which is generated in a previous cycle during repeated execution and a second change which is an amount of change of a direction with respect to a direction directed from the vehicle to a point a predetermined distance away on the target trajectory between the previous cycle and a current cycle are decreased. SUMMARY In view of the above-mentioned shortcomings of the prior art, a purpose of the present application is to solve the problem that a self-driving vehicle cannot plan a path when a travelling space is narrow, and cannot effectively avoid multiple obstacles. The invention is set out in the appended set of claims. In order to solve the above technical problem, the present invention provides a path planning method, a path planning apparatus, an electronic device and a storage medium. Solutions are as follows. In one aspect, a path planning method is provided, including: projecting information on lane lines, information on a lane center-line and information on an obstacle contour into a target coordinate system to obtain projection information;determining a start point and an end point of path planning in the target coordinate system, and if it is determined based on the projection information that there is an obstacle between the start point and the end point, then performing sampling on both sides of the obstacle contour along a travelling direction of a vehicle to obtain sampling points;determining a plurality of paths based on the start point, the end point and the sampling points;performing a cost calculation on the plurality of paths, and determining a reference path in the plurality of paths based on results of the cost calculation, where the results of the cost calculation characterize degrees to which the paths are close to the lane center-line in a case of avoiding the obstacle; anddetermining a target planned path based on the reference path. Optionally, performing the sampling on both sides of the obstacle contour along the travelling direction of the ve