Cost design for path selection in autonomous driving technology

Patent metadata

Field	Value
Publication number	US-10824153-B2
Application number	US-201815954352-A
Country	US
Kind code	B2
Filing date	Apr 16, 2018
Priority date	Apr 16, 2018
Publication date	Nov 3, 2020
Grant date	Nov 3, 2020

Abstract

A new cost design is disclosed for evaluating candidate path curves for navigating an autonomous driving vehicle (ADV) through a segment of a route which may include an obstacle. Each point on each candidate path curve has a plurality of attributes having logical values and an associated priority of evaluation, and at least one numeric attribute having an associated priority of evaluation. A cost for each path curve is determined using the attributes and priorities, and a least cost path curve is selected using the attributes and priorities. By comparing attribute values in accordance with priority, and utilizing logical values, the efficiency of determining path curve cost and selecting a least cost path curve is substantially improved.

First claim

What is claimed is:

1. A computer-implemented method for generating a path to drive an autonomous driving vehicle, the method comprising:
generating a plurality of paths as path candidates for an autonomous driving vehicle (ADV) to navigate through a lane in view of an obstacle, each path having a plurality of points, wherein each point is associated with a plurality of attributes corresponding to a plurality of categories, each category being associated with a different priority;
for each of the plurality of paths,
examining each of the points of the path in view of the obstacle and a lane configuration of the lane to determine whether the point of the path satisfies each of the predetermined categories,
assigning a cost value to each of the attributes of the point of the path based on whether the point of the path satisfies a corresponding category of each attribute; and

traversing the attributes of the paths from higher priorities to lower priorities to compare and select a path with a least total cost based on the cost values of the attributes, including removing at least one path from consideration if there is a difference between cost values of attributes of multiple paths with an identical priority; and
navigating the ADV using the selected path with the least total cost.

2. The method of claim 1 , further comprising, for each attribute of a particular category of each path, aggregating the cost values of the attributes of the same category of all points of the path to generate an aggregated cost value for the attribute, wherein comparing and selecting a path are performed based on the aggregated cost values of the attributes.

3. The method of claim 1 , wherein the plurality of categories comprise:
an obstacle category associated with a first priority to indicate whether the ADV potentially collides with the obstacle; and
a lane boundary category associated with a second priority to indicate whether the ADV potentially move outside of a lane boundary of the lane, wherein the first priority is higher than the second priority.

4. The method of claim 3 , wherein an attribute corresponding to the obstacle category or the lane boundary category is a binary attribute having a predetermined logical value indicating whether the attribute satisfies a corresponding category.

5. The method of claim 3 , wherein the plurality of categories further comprise a comfort category associated with a third priority representing a comfort level of a passenger of the ADV, wherein the third priority is lower than the second priority.

6. The method of claim 5 , wherein an attribute corresponding to the comfort category contains a numeric value representing the comfort level determined based in part on a curvature of a corresponding path.

7. The method of claim 3 , wherein the lane boundary category comprises:
a physical lane boundary associated with a fourth priority, the physical lane boundary representing a boundary of a physical lane; and
a virtual lane boundary associated with a fifth priority, the virtual lane boundary representing a boundary of one of a plurality of virtual lanes of a physical lane, wherein the fourth priority is higher than the fifth priority.

8. A non-transitory computer-readable programmed with executable instructions that, when executed by a processing system comprising at least one hardware processor, perform operations that implement a method for determining a least cost path curve for a plurality of path curves for operating an autonomous driving vehicle (ADV), the operations comprising:
generating a plurality of paths as path candidates for an autonomous driving vehicle (ADV) to navigate through a lane in view of an obstacle, each path having a plurality of points, wherein each point is associated with a plurality of attributes corresponding to a plurality of categories, each category being associated with a different priority;
for each of the plurality of paths,
examining each of the points of the path in view of the obstacle and a lane configuration of the lane to determine whether the point of the path satisfies each of the predetermined categories,
assigning a cost value to each of the attributes of the point of the path based on whether the point of the path satisfies a corresponding category of each attribute; and

traversing the attributes of the paths from higher priorities to lower priorities to compare and select a path with a least total cost based on the cost values of the attributes, including removing at least one path from consideration if there is a difference between cost values of attributes of multiple paths with an identical priority; and
navigating the ADV using the selected path with the least total cost.

9. The medium of claim 8 , the operations further comprising, for each attribute of a particular category of each path, aggregating the cost values of the attributes of the same category of all points of the path to generate an aggregated cost value for the attribute, wherein comparing and selecting a path are performed based on the aggregated cost values of the attributes.

10. The medium of claim 8 , wherein the plurality of categories comprise:
an obstacle category associated with a first priority to indicate whether the ADV potentially collides with the obstacle; and
a lane boundary category associated with a second priority to indicate whether the ADV potentially move outside of a lane boundary of the lane, wherein the first priority is higher than the second priority.

11. The medium of claim 10 , wherein an attribute corresponding to the obstacle category or the lane boundary category is a binary attribute having a predetermined logical value indicating whether the attribute satisfies a corresponding category.

12. The medium of claim 10 , wherein the plurality of categories further comprise a comfort category associated with a third priority representing a comfort level of a passenger of the ADV, wherein the third priority is lower than the second priority.

13. The medium of claim 12 , wherein an attribute corresponding to the comfort category contains a numeric value representing the comfort level determined based in part on a curvature of a corresponding path.

14. The medium of claim 10 , wherein the lane boundary category comprises:
a physical lane boundary associated with a fourth priority, the physical lane boundary representing a boundary of a physical lane; and
a virtual lane boundary associated with a fifth priority, the virtual lane boundary representing a boundary of one of a plurality of virtual lanes of a physical lane, wherein the fourth priority is higher than the fifth priority.

15. A system comprising a processing system comprising at least one hardware processor, coupled to a memory programmed with executable instructions that when executed by the processing system, perform operations that implement a method for determining a least cost path curve for a plurality of path curves for operating an autonomous driving vehicle (ADV), the operations comprising:
generating a plurality of paths as path candidates for an autonomous driving vehicle (ADV) to navigate through a lane in view of an obstacle, each path having a plurality of points, wherein each point is associated with a plurality of attributes corresponding to a plurality of categories, each category being associated with a different priority;
for each of the plurality of paths,
examining each of the points of the path in view of the obstacle and a lane configuration of the lane to determine whether the point of the path satisfies each of the predetermined categories, and
assigning a cos

Assignees

• Baidu Usa Llc

Inventors

• Zhang Liangliang
• Li Dong
• Hu Jiangtao
• Tao Jiaming
• Jiang Yifei

Classifications

• G08G1/165 Primary

  for passive traffic, e.g. including static obstacles, trees · CPC title

• G08G1/096844

  where the complete route is dynamically recomputed based on new data · CPC title

• G08G1/096725

  where the received information generates an automatic action on the vehicle control · CPC title

• G01C21/3658

  Lane guidance · CPC title

• G08G1/096775

  where the origin of the information is a central station · CPC title