Offline optimization to robot behavior

Patent metadata

Field	Value
Publication number	US-11084167-B1
Application number	US-201816108919-A
Country	US
Kind code	B1
Filing date	Aug 22, 2018
Priority date	Dec 29, 2014
Publication date	Aug 10, 2021
Grant date	Aug 10, 2021

Abstract

A computing system may provide a model of a robot. The model may be configured to determine simulated motions of the robot based on sets of control parameters. The computing system may also operate the model with multiple sets of control parameters to simulate respective motions of the robot. The computing system may further determine respective scores for each respective simulated motion of the robot, wherein the respective scores are based on constraints associated with each limb of the robot and a predetermined goal. The constraints include actuator constraints and joint constraints for limbs of the robot. Additionally, the computing system may select, based on the respective scores, a set of control parameters associated with a particular score. Further, the computing system may modify a behavior of the robot based on the selected set of control parameters to perform a coordinated exertion of forces by actuators of the robot.

First claim

What is claimed is:

1. A method comprising:
simulating, by a computing device of a robot comprising a limb including a joint connecting a first member and a second member, a first motion of the robot with a first set of control parameters using a model, the first set of control parameters representing a first exertion of force by an actuator of the robot, the actuator configured to control:
a state of the limb, wherein the state of the limb includes an angle between the first member and the second member; and
an extent of torque exerted at the joint, wherein the torque exerted at the joint results from an exertion of a force by the actuator;

determining, by the computing device, a second set of control parameters based on the first set of control parameters and the simulated first motion of the robot, the second set of control parameters representing a second exertion of force by the actuator of the robot;
simulating, by the computing device, a second motion of the robot with the second set of control parameters using the model;
determining, by the computing device, respective first and second scores for the first and second simulated motions of the robot, wherein the first and second scores are based on constraints associated with at least one limb of the robot and a predetermined goal;
selecting, by the computing device, one of the first set of control parameters or the second set of control parameters based on the first and second scores, the selected one of the first set of control parameters or the second set of control parameters including a starting state for the actuator, an ending state for the actuator, and a duration of time for the associated one of the first exertion of force or the second exertion force by the actuator; and
actuating, by the computing device, the actuator to exert the one of the first exertion of force or the second exertion of force that is associated with the selected one of the first set of control parameters or the second set of control parameters.

2. The method of claim 1 , wherein the joint is constrained to operate within an operable range of angles indicated by a joint constraint, and wherein the actuator is constrained to operate within an operable range of forces indicated by an actuator constraint.

3. The method of claim 1 , wherein simulating either one of the first motion of the robot or the second motion of the robot includes the joint taking on various angles and torques over the duration of time.

4. The method of claim 3 , wherein constraints associated with the at least one limb of the robot include actuator constraints of the actuator and joint constraints of the joint, wherein the model is further configured to include a representation of a control system of the robot, and wherein simulating the first motion of the robot with the first set of control parameters using the model comprises:
providing the first set of control parameters to the representation of the control system; and
based on the model and the first set of control parameters, simulating the first motion of the robot limited by the respective actuator constraints and joint constraints.

5. The method of claim 1 , wherein actuating the at least one actuator to exert the one of the first exertion of force or the second exertion of force comprises:
transitioning from a first mode of operation to a second mode of operation, the first mode of operation comprising a steady-state mode of operation that balances the robot and the second mode of operation comprising a transient mode of operation that causes the robot to:
operate the actuator in the starting state; and
transition the actuator from the starting state to the ending state over the duration of time; and

upon completion of the second mode of operation, transitioning from the second mode of operation to the first mode of operation.

6. The method of claim 1 , wherein determining the respective first score for the simulated first motion of the robot comprises:
determining a first sub-score based on a relationship between a first measurable aspect of the simulated first motion and the predetermined goal;
determining a second sub-score based on a relationship between a second measurable aspect of the simulated first motion and the constraints associated with the at least one limb of the robot; and
calculating the first score based on the first sub-score and the second sub-score.

7. The method of claim 1 , wherein determining the second set of control parameters comprises:
identifying a pattern in the simulated first motion that indicates a manner of varying the first set of control parameters that results in a further simulated motion associated with a further score that approaches a desired value, wherein the desired value numerically represents the predetermined goal; and
determining the second set of control parameters based on the first set of control parameters and the identified pattern.

8. The method of claim 1 , wherein selecting one of the first set of control parameters or the second set of control parameters comprises:
determining a particular score from the first and second scores that is nearest to a desired value, wherein the desired value numerically represents the predetermined goal; and
selecting the one of the first set of control parameters or the second set of control parameters that corresponds to the particular score.

9. The method of claim 1 , wherein the selected one of the first set of control parameters or the second set of parameters includes a plurality of force values, and wherein actuating the actuator to exert the one of the first exertion of force or the second exertion of force comprises:
exerting a first magnitude of force specified by an initial force value from the plurality of force values of the selected one of the first set of control parameters or the second set of control parameters; and
exerting a second magnitude of force specified by a subsequent force value from the plurality of force values of the selected one of the first set of control parameters or the second set of control parameters.

10. The method of claim 1 , wherein the selected one of the first set of control parameters or the second set of control parameters defines a transient motion of the robot that exceeds a constraint of the actuator, the method further comprising:
selecting a third set of control parameters that define a recovery motion to maintain balance of the robot, wherein the recovery motion includes a third exertion of force by the actuator of the robot; and
actuating the actuator to exert the third exertion of force.

11. A robot comprising:
at least one limb including a joint connecting a first member and a second member;
an actuator configured to control:
a state of the limb, wherein the state of the limb includes an angle between the first member and the second member; and
an extent of torque exerted at the joint, wherein the torque exerted at the joint results from an exertion of a force by the actuator;

a processor; and
a memory storing a model of the robot and program logic, the program logic executable by the processor to cause the robot to perform operations comprising:
simulating a first motion of the robot with a first set of control parameters using a model, the first set of control parameters representing a first exertion of force by an actuator of the robot;
determining a second set of control parameters based on the first set of control parameters and the simulated first motion of the robot, the second set of control parameters representing a second exertion of force by the actuator of the robot;
simulating a second motion of the robot with the second set of contro

Assignees

• Boston Dynamics Inc

Inventors

• Silva Marco Da
• Stephens Benjamin
• Rizzi Alfred Anthony
• Abe Yeuhi

Classifications

• Y10S901/02

  Arm motion controller · CPC title

• B25J9/1605 Primary

  Simulation of manipulator lay-out, design, modelling of manipulator · CPC title

• Y10S901/01

  Mobile robot · CPC title

• B62D57/032

  with alternately or sequentially lifted supporting base and legs; with alternately or sequentially lifted feet or skid (B62D57/024 takes precedence) · CPC title

• Y10S901/28

  Joint · CPC title