Method for autonomously driving a vehicle based on moving trails of obstacles surrounding the vehicle

Patent metadata

Field	Value
Publication number	US-11679764-B2
Application number	US-201916457847-A
Country	US
Kind code	B2
Filing date	Jun 28, 2019
Priority date	Jun 28, 2019
Publication date	Jun 20, 2023
Grant date	Jun 20, 2023

Abstract

During the autonomous driving, the movement trails or moving history of obstacles, as well as, an autonomous driving vehicle (ADV) may be maintained in a corresponding buffer. For the obstacles and the ADV, the vehicle states at different points in time are maintained and stored in one or more buffers. The vehicle states representing the moving trails or moving history of the obstacles and the ADV may be utilized to reconstruct a history trajectory of the obstacles and the ADV, which may be used for a variety of purposes. For example, the moving trails or history of obstacles may be utilized to determine lane configuration of one or more lanes of a road, particularly, in a rural area where the lane markings are unclear. The moving history of the obstacles may also be utilized predict the future movement of the obstacles, tailgate an obstacle, and infer a lane line.

First claim

What is claimed is:

1. A computer-implemented method for operating an autonomous driving vehicle (ADV), the method comprising:
perceiving a driving environment surrounding an ADV based on sensor data obtained from a plurality of sensors, including detecting one or more moving obstacles;
allocating one or more obstacle state buffers, one for each of the one or more moving obstacles to store obstacle states of a corresponding moving obstacle at different points in time for a predetermined time period, wherein each obstacle state includes at least a location of the corresponding moving obstacle at a particular point in time;
inferring a lane center line of a lane based on the obstacle states retrieved from at least one of the obstacle state buffers;
determining a lane width based on the inferred lane center line, an obstacle width of a corresponding moving obstacle, and a required minimum clearance space;
predicting a further movement of at least one of the one or more moving obstacles based on past obstacle states stored in an associated obstacle state buffer; and
planning a trajectory to drive the ADV based on the lane width and the predicted further movement of the at least one of the one or more moving obstacles to avoid a collision with any one of the one or more moving obstacles.

2. The method of claim 1 , wherein movement of the at least one of the one or more moving obstacles is not able to be sensed by the ADV and would cause the one of the one or more moving obstacles to collide with the ADV if the trajectory was not based on the predicted further movement.

3. The method of claim 1 , wherein predicting the further movement of the at least one of the one or more moving obstacles comprises:
reconstructing a first past moving trajectory for the at least one of the one or more moving obstacles for a first past period of time with a first portion of the past obstacle states; and
predicting a second past moving trajectory for the at least one of the one or more moving obstacles for a second past period of time using a past moving trajectory.

4. The method of claim 1 , further comprising performing an analysis on the obstacle states stored in the obstacle state buffers to determine lane configuration of the driving environment without using map data of a map associated with the driving environment.

5. The method of claim 4 , further comprising reconstructing a moving trajectory of each of the one or more moving obstacles based on its obstacle states retrieved from the corresponding obstacle state buffer, wherein the lane configuration is derived based on the moving trajectories of the one or more moving obstacles.

6. The method of claim 1 , further comprising:
detecting that a first moving obstacle is blocked by a static obstacle; and
predicting movement of the first moving obstacle based on obstacle states stored in a first obstacle state buffer associated with the first moving obstacle, while the first moving obstacle remains blocked by the static obstacle.

7. The method of claim 1 , further comprising:
in response to a request for following a second moving obstacle, retrieving obstacle states from a second obstacle state buffer associated with the second moving obstacle;
reconstructing a second moving trajectory from the obstacle states retrieved from the second obstacle state buffer; and
planning a trajectory for the ADV based on the second moving trajectory.

8. A non-transitory machine-readable medium having instructions stored therein, which when executed by a processor, cause the processor to perform operations, the operations comprising:
perceiving a driving environment surrounding an ADV based on sensor data obtained from a plurality of sensors, including detecting one or more moving obstacles;
allocating one or more obstacle state buffers, one for each of the one or more moving obstacles to store obstacle states of a corresponding moving obstacle at different points in time for a predetermined time period, wherein each obstacle state includes at least a location of the corresponding moving obstacle at a particular point in time;
inferring a lane center line of a lane based on the obstacle states retrieved from at least one of the obstacle state buffers;
determining a lane width based on the inferred lane center line, an obstacle width of a corresponding moving obstacle, and a required minimum clearance space;
predicting a further movement of at least one of the one or more moving obstacles based on past obstacle states stored in an associated obstacle state buffer; and
planning a trajectory to drive the ADV based on the lane width and the predicted further movement of the at least one of the one or more moving obstacles to avoid a collision with any one of the one or more moving obstacles.

9. The machine-readable medium of claim 8 , wherein movement of the at least one of the one or more moving obstacles is not able to be sensed by the ADV and would cause the one of the one or more moving obstacles to collide with the ADV if the trajectory was not based on the predicted further movement.

10. The machine-readable medium of claim 8 , wherein predicting the further movement of the at least one of the one or more moving obstacles comprises:
reconstructing a first past moving trajectory for the at least one of the one or more moving obstacles for a first past period of time with a first portion of the past obstacle states; and
predicting a second past moving trajectory for the at least one of the one or more moving obstacles for a second past period of time using a past moving trajectory.

11. The machine-readable medium of claim 8 , wherein the operations further comprise performing an analysis on the obstacle states stored in the obstacle state buffers to determine lane configuration of the driving environment without using map data of a map associated with the driving environment.

12. The machine-readable medium of claim 11 , wherein the operations further comprise reconstructing a moving trajectory of each of the one or more moving obstacles based on its obstacle states retrieved from the corresponding obstacle state buffer, wherein the lane configuration is derived based on the moving trajectories of the one or more moving obstacles.

13. The machine-readable medium of claim 8 , wherein the operations further comprise:
detecting that a first moving obstacle is blocked by a static obstacle; and
predicting movement of the first moving obstacle based on obstacle states stored in a first obstacle state buffer associated with the first moving obstacle, while the first moving obstacle remains blocked by the static obstacle.

14. The machine-readable medium of claim 8 , wherein the operations further comprise:
in response to a request for following a second moving obstacle, retrieving obstacle states from a second obstacle state buffer associated with the second moving obstacle;
reconstructing a second moving trajectory from the obstacle states retrieved from the second obstacle state buffer; and
planning a trajectory for the ADV based on the second moving trajectory.

15. A data processing system, comprising:
a processor; and
a memory coupled to the processor to store instructions, which when executed by the processor, cause the processor to perform operations, the operations including:
perceiving a driving environment surrounding an ADV based on sensor data obtained from a plurality of sensors, including detecting one or more moving obstacles;
allocating one or more obstacle state buffers, one for each of the one or more moving obstacles to store obstacle states of a corresponding m

Assignees

• Baidu Usa Llc

Inventors

• Choe Tae Eun
• Chen Guang
• Zhang Weide
• Guo Yuliang
• Tsoi Ka Wai

Classifications

• B60W30/0956 Primary

  the prediction being responsive to traffic or environmental parameters · CPC title

• G05D1/0088

  characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours (using knowledge based models G06N5/00) · CPC title

• B60W30/0953

  the prediction being responsive to vehicle dynamic parameters · CPC title

• B60W30/09 Primary

  Taking automatic action to avoid collision, e.g. braking and steering · CPC title

• G06V20/588

  Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road · CPC title