Processor-implemented systems and methods for automated driving

• US 10,146,224 B2
• Filed: 11/09/2016
• Issued: 12/04/2018
• Est. Priority Date: 11/09/2016
• Status: Active Grant

• Current Assignee: GM Global Technology Operations LLC (General Motors Company)
• Original Assignee: GM Global Technology Operations LLC (General Motors Company)
• Inventors: Tafti, Sayyed Rouhollah Jafari, Zeng, Shuqing
• Primary Examiner(s): Beaulieu, Yonel
• Application Number:
  US15/347,546
• Publication Number:
  US 20180129203A1
• Time in Patent Office: 755 Days
• CPC Class Codes: B60W 60/001 Planning or execution of driving tasks
• CPC Class Codes: G01C 21/3453 Special cost functions, i.e. other than distance or default speed limit of road segments
• CPC Class Codes: G05D 1/0005 with arrangements to save energy
• CPC Class Codes: G05D 1/0088 characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours using knowledge based models G06N5/00
• CPC Class Codes: G05D 1/0212 with means for defining a desired trajectory involving a plurality of land vehicles G05D1/0287
• CPC Class Codes: G05D 1/0217 in accordance with energy consumption, time reduction or distance reduction criteria
• CPC Class Codes: G05D 1/0278 using satellite positioning signals, e.g. GPS

Abstract

Systems and methods are provided for controlling a vehicle. In one embodiment, a method includes receiving vehicle and object environment data. A search graph is generated based upon the received data. The search graph contains a grid of points for locating objects and is used to determine a desired trajectory for the vehicle.

Claims

1. A processor-implemented method for automated driving of a vehicle, comprising:

receiving, by one or more data processors, vehicle state data, driving scenario plan, behavior control data and vehicle object environment data;

generating, by the one or more data processors, a search graph based upon the received vehicle state data and driving scenario plan;

wherein the search graph contains a grid of points for locating objects relative to a reference trajectory path;

wherein the search graph uses a coordinate system that indicates lateral offsets of the objects relative to the reference trajectory path stations; and

determining, by the one or more data processors, a trajectory path that is optimal with respect to a specific cost function;

wherein the optimal trajectory path is provided to a controller in the vehicle for implementing the driving scenario plan.

2. The method of claim 1, wherein the objects include other vehicles on the road.

3. The method of claim 1, wherein the determining of the trajectory path includes using a cost function during optimization of the trajectory path, wherein the cost function contains factors reactive to changes in dynamics of the vehicle and the objects on the road.

4. The method of claim 3, where the cost function includes offset from the reference trajectory path, steering action cost, offsets from a previously planned trajectory path, and a collision cost factor for the vehicle with respect to one or more of the objects.

5. The method of claim 1, wherein search graphs are generated using the coordinate system to determine trajectory paths in different road scenarios.

6. The method of claim 5, wherein the road scenarios include sharp turns, U-turns, sharp turns in urban settings, and high-speed driving for highway roads.

7. The method of claim 1 further comprising:

smoothing the optimal trajectory path based upon minimizing curvature and curvature rates within a safe corridor associated with the search graph;

wherein the safe corridor defines a distance range from the optimal path within which the vehicle is at a safe distance from the objects.

8. The method of claim 7 further comprising:

generating a local trajectory based upon the smoothed trajectory path;

wherein the local trajectory is determined based upon satisfying continuity conditions for reaching the destination points on the optimal trajectory at different lookahead times.

9. The method of claim 8, wherein the local trajectory satisfies kinematic and dynamic constraints of the vehicle and free-space and road boundary constraints of the environment and is directly used by a longitudinal and lateral controller in the vehicle.

10. The method of claim 8, wherein the local trajectory follows the rules of traffic, is safe within the road boundaries, and considers vehicle non-holonomic motion constraints and dynamic constraints.

11. The method of claim 1, wherein the received vehicle state data, driving scenario plan, and vehicle object environment data are provided by one of the following modules:

an object fusion module for listing detected remote road users and obstacles on the road;

a scenario plan module which includes road boundaries using map data and lane fusion along with a reference trajectory assuming obstacle-free environments;

a behavior control module for handling behavioral instructions with consideration of traffic participants and scenarios; and

a vehicle state module for determining current state of the vehicle.

12. The method of claim 1, wherein the search graph includes nodes which represent points along the reference trajectory path stations with different lateral offsets.

13. The method of claim 1, wherein the search graph includes line segments connecting the nodes to neighboring nodes.

14. The method of claim 13, wherein each node has the same velocity as its corresponding reference trajectory point.

15. The method of claim 13, wherein the nodes are associated with velocity profiles relative to the reference trajectory path.

16. The method of claim 15, wherein the velocity profiles are determined based on the objects'"'"' speed on the road, acceleration and deceleration limits, and road speed limit.

17. A processor-implemented system for automated driving of a vehicle, comprising:

a data communication pathway for transmitting vehicle state data, driving scenario plan, and vehicle object environment data;

one or more data processors configured to;

generate a search graph based upon the received vehicle state data and driving scenario plan;

wherein the search graph contains a grid of points for locating objects relative to a reference trajectory path;

wherein the search graph uses a coordinate system that indicates lateral offsets of the objects relative to the reference trajectory path stations; and

determine a trajectory path that is optimal with respect to a specific cost function;

wherein the optimal trajectory path is provided to a controller in the vehicle for implementing the driving scenario plan.

18. An autonomous vehicle, comprising:

at least one sensor that provides sensor data about objects within the vehicle'"'"'s environment and about vehicle state; and

a controller that, by a processor and based on the sensor data, is configured to;

generate a search graph based upon the vehicle state data and a driving scenario plan;

wherein the search graph contains a grid of points for locating objects relative to a reference trajectory path;

wherein the search graph uses a coordinate system that indicates lateral offsets of the objects relative to the reference trajectory path; and

determine a trajectory path that is optimal with respect to a specific cost function;

wherein the optimal trajectory path is provided to a controller in the vehicle for implementing the driving scenario plan.