System and method for providing multiple agents for decision making, trajectory planning, and control for autonomous vehicles

Patent metadata

Field	Value
Publication number	US-12242271-B2
Application number	US-202217983974-A
Country	US
Kind code	B2
Filing date	Nov 9, 2022
Priority date	Sep 30, 2017
Publication date	Mar 4, 2025
Grant date	Mar 4, 2025

Abstract

A system and method for providing multiple agents for decision making, trajectory planning, and control for autonomous vehicles are disclosed. A particular embodiment includes: partitioning a multiple agent autonomous vehicle control module for an autonomous vehicle into a plurality of subsystem agents, the plurality of subsystem agents including a deep computing vehicle control subsystem and a fast response vehicle control subsystem; receiving a task request from a vehicle subsystem; determining if the task request is appropriate for the deep computing vehicle control subsystem or the fast response vehicle control subsystem based on content of the task request or a context of the autonomous vehicle; dispatching the task request to the deep computing vehicle control subsystem or the fast response vehicle control subsystem based on the determination; causing execution of the deep computing vehicle control subsystem or the fast response vehicle control subsystem by use of a data processor to produce a vehicle control output; and providing the vehicle control output to a vehicle control subsystem of the autonomous vehicle.

First claim

What is claimed is:

1. A system comprising:
a data processor;
a memory for storing modules, executable by the data processor, the modules being data processing modules configured to perform autonomous vehicle control operations for an autonomous vehicle;
a decision making subsystem executable by the data processor, the decision making subsystem including a decision making deep computing agent and a decision making fast response agent;
a trajectory generation subsystem executable by the data processor, the trajectory generation subsystem including a trajectory generation deep computing agent and a trajectory generation fast response agent;
a deep computing vehicle control module executable by the data processor, the deep computing vehicle control module processing task requests from the decision making deep computing agent and the trajectory generation deep computing agent to produce an autonomous vehicle motion plan; and
a fast response vehicle control module executable by the data processor, the fast response vehicle control module processing task requests from the decision making fast response agent and the trajectory generation fast response agent to produce the autonomous vehicle motion plan, the fast response vehicle control module configured to preempt the deep computing vehicle control module;
the system being configured to use the autonomous vehicle motion plan to control movement of the autonomous vehicle.

2. The system of claim 1 being further configured to receive sensor data from a plurality of sensors on the autonomous vehicle.

3. The system of claim 1 wherein the fast response vehicle control module is configured to cause modification of parameters used by the deep computing vehicle control module.

4. The system of claim 1 wherein the execution of the decision making subsystem and the trajectory generation subsystem is performed by different data processors.

5. The system of claim 1 wherein the execution of the deep computing vehicle control module and the fast response vehicle control module is performed by different data processors.

6. The system of claim 1 wherein the execution of the decision making deep computing agent and the decision making fast response agent is performed by different data processors.

7. The system of claim 1 wherein the execution of the trajectory generation deep computing agent and the trajectory generation fast response agent is performed by different data processors.

8. The system of claim 1 wherein execution of the decision making subsystem and the trajectory generation subsystem by use of the data processor is performed in parallel.

9. The system of claim 1 wherein the autonomous vehicle motion plan is provided to a vehicle control subsystem causing the autonomous vehicle to follow a routing and perform vehicle motion control operations corresponding to the autonomous vehicle motion plan.

10. A method in an autonomous vehicle control system to more efficiently and safely control an autonomous vehicle with fast and efficient control data processing, the method comprising:
partitioning a multiple agent autonomous vehicle control module for an autonomous vehicle into a plurality of subsystems and subsystem agents, the plurality of subsystems and subsystem agents being data processing modules executable by a data processor and configured to perform autonomous vehicle control operations for an autonomous vehicle, the plurality of subsystems and subsystem agents including: a decision making subsystem executable by the data processor, the decision making subsystem including a decision making deep computing agent and a decision making fast response agent; and a trajectory generation subsystem executable by the data processor, the trajectory generation subsystem including a trajectory generation deep computing agent and a trajectory generation fast response agent;
executing, by use of the data processor, a deep computing vehicle control module to process task requests from the decision making deep computing agent and the trajectory generation deep computing agent to produce an autonomous vehicle motion plan; and
executing, by use of the data processor, a fast response vehicle control module to process task requests from the decision making fast response agent and the trajectory generation fast response agent to produce the autonomous vehicle motion plan, the fast response vehicle control module configured to preempt the deep computing vehicle control module; and
using the autonomous vehicle motion plan to control movement of the autonomous vehicle.

11. The method of claim 10 further including receiving sensor data from a plurality of sensors on the autonomous vehicle.

12. The method of claim 10 wherein the fast response vehicle control module is configured to cause modification of parameters used by the deep computing vehicle control module.

13. The method of claim 10 wherein preemption of the deep computing vehicle control module comprises use of a pre-emptive multitasking operating system.

14. The method of claim 10 wherein the execution of the decision making subsystem and the trajectory generation subsystem is performed by different data processors.

15. The method of claim 10 wherein the execution of the deep computing vehicle control module and the fast response vehicle control module is performed by different data processors.

16. The method of claim 10 wherein the execution of the decision making deep computing agent and the decision making fast response agent is performed by different data processors.

17. The method of claim 10 wherein the execution of the trajectory generation deep computing agent and the trajectory generation fast response agent is performed by different data processors.

18. The method of claim 10 wherein the autonomous vehicle motion plan is used by the vehicle control system to cause the autonomous vehicle to follow a routing and perform vehicle motion control operations corresponding to the autonomous vehicle motion plan.

19. A non-transitory machine-useable storage medium embodying instructions in an autonomous vehicle control system to more efficiently and safely control an autonomous vehicle with fast and efficient control data processing, the instructions being partitioned into plurality of subsystems and subsystem agents being executable by a data processor, the plurality of subsystems and subsystem agents including: a decision making subsystem executable by the data processor, the decision making subsystem including a decision making deep computing agent and a decision making fast response agent; and a trajectory generation subsystem executable by the data processor, the trajectory generation subsystem including a trajectory generation deep computing agent and a trajectory generation fast response agent, the instructions being further configured to:
execute, by use of the data processor, a deep computing vehicle control module to process task requests from the trajectory generation deep computing agent and the vehicle control deep computing agent to produce a vehicle control output;
execute, by use of the data processor, a fast response vehicle control module to process task requests from the trajectory generation fast response agent and the vehicle control fast response agent to produce the vehicle control output, the fast response vehicle control module configured to preempt the deep computing vehicle control module; and
use the vehicle control output to control movement of the autonomous vehicle.

20. The non-transitory machine-

Assignees

• Tusimple Inc

Inventors

• Sun Xing
• Zhao Yufei
• Lin Wutu
• Xuan Zijie
• Liu Liu
• Ma Kai-Chieh

Classifications

• G05D1/20

  Control system inputs · CPC title

• G05D2101/10

  using artificial intelligence [AI] techniques · CPC title

• G05D1/81

  Handing over between on-board automatic and on-board manual control · CPC title

• B60W30/00

  Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units · CPC title

• G06F9/00

  Arrangements for program control, e.g. control units (program control for peripheral devices G06F13/10) · CPC title