Determination of foot placement for humanoid push recovery

• US 7,949,430 B2
• Filed: 11/28/2007
• Issued: 05/24/2011
• Est. Priority Date: 11/29/2006
• Status: Expired due to Fees

• Current Assignee: Florida Institute For Human and Machine Cognition, Honda Motor Co., Ltd. (Honda Motor Company)
• Original Assignee: Florida Institute For Human and Machine Cognition, Honda Motor Co., Ltd. (Honda Motor Company)
• Inventors: Pratt, Jerry, Goswami, Ambarish
• Primary Examiner(s): Black; Thomas G
• Assistant Examiner(s): Olsen; Lin B
• Application Number:
  US11/946,702
• Publication Number:
  US 20080133053A1
• Time in Patent Office: 1,273 Days
• Field of Search: 901/27, 901/30, 700/252
• US Class Current: 700/253
• CPC Class Codes: B62D 57/032 with alternately or sequentially lifted supporting base and legs; with alternately or sequentially lifted feet or skid B62D57/024 takes precedence

Abstract

A legged robot subjected to a force is controlled by determining an instantaneous capture point where the robot will step with a swing leg to reach a balanced home position, the balanced home position being a state in which the Center of Mass remains substantially over the Center of Pressure and the robot is able to maintain its balance indefinitely. The capture point can be determined using a Linear Inverted Pendulum Plus Flywheel (LIPPF) model of the robot. The LIPPF model includes a flywheel with a mass and a rotational inertia, and a variable length leg link. A torque profile is applied to the flywheel and a set of capture points is determined based on this torque profile An experimentally determined error value can be added to a capture point that is determined based on the model to account for differences between an actual robot and the model.

Claims

1. A method for controlling a robot having at least two legs including a swing leg, the robot standing on a surface and subjected to a force, the method comprising:

receiving, by a controller, a state vector comprising positions and velocities of joints of the robot;

applying, by the controller, a linear motion model to the state vector that models at least a portion of the robot as an inverted pendulum, the linear motion model when applied to the state vector determining an instantaneous capture point on the surface where the robot will step with the swing leg to reach a balanced home position, the balanced home position being a state in which a center of mass of the robot remains substantially over a center of pressure of the robot such that the robot maintains balance; and

outputting a control signal to an actuator that causes the robot to step with the swing leg to the instantaneous capture point.

2. The method of claim 1, wherein the instantaneous capture point is determined such that if the robot steps to the instantaneous capture point, a linear inverted pendulum orbital energy, E_LIP, will equal zero.

3. The method of claim 1, wherein the instantaneous capture point is located at a horizontal distance x_capture from the center of mass of the robot, where

4. The method of claim 1, wherein the robot is a biped.

5. The method of claim 1, wherein the balanced home position further comprises a state in which the robot is able to maintain its balance on the swing leg indefinitely without taking another step.

6. The method of claim 1, wherein the balanced home position is a complete stop.

7. A method for controlling a robot having at least two legs including a swing leg, the robot standing on a surface and subjected to a force, the method comprising:

receiving, by a controller, a state vector comprising instantaneous positions and velocities of joints of the robot;

applying, by the controller, a linear motion model to the state vector that models a first portion of the robot as an inverted pendulum and models a second portion of the robot as a flywheel, the linear motion model when applied to the state vector determining a capture region on the surface where the robot will step with the swing leg to reach a balanced home position, the balanced home position being a state in which a center of mass of the robot remains substantially over a center of pressure of the robot such that the robot maintains balance; and

outputting a first control signal to an actuator that causes the robot to step with the swing leg to a point within the capture region.

8. The method of claim 7, further comprising:

determining an angular momentum that the robot will generate to reach the balanced home position; and

outputting a second control signal to the actuator that causes the robot to move according to the determine angular momentum.

9. The method of claim 7, wherein the linear motion model comprises a linear inverted pendulum plus flywheel model of the robot, the model comprising:

a flywheel with a mass and a rotational inertia, and

a variable length leg link connected to the flywheel.

10. The method of claim 9, wherein the flywheel is capable of instantaneous velocity change.

11. The method of claim 10, wherein the boundaries of the capture region are determined at least in part by a maximum velocity change of the flywheel and a minimum velocity change of the flywheel.

12. The method of claim 9, wherein the flywheel is capable of instantaneous position change.

13. The method of claim 12, wherein boundaries of the capture region are determined at least in part by a maximum position change of the flywheel and a minimum position change of the flywheel.

14. The method of claim 9, wherein a torque profile is applied to the flywheel, the torque profile comprising accelerating the flywheel in one direction for a period of time, decelerating the flywheel for a period of time, and stopping the flywheel at a maximum flywheel angle.

15. The method of claim 14, wherein boundaries of the capture region are determined at least in part by a maximum torque applied to the flywheel and a minimum torque applied to the flywheel.

16. The method of claim 9, wherein the capture region is a horizontal distance range from the center of mass of the robot, the horizontal distance range being greater than or equal to an inner boundary x_capture1 and less than or equal to an outer boundary x_capture2, wherein the inner boundary x_capture1 comprises a minimum horizontal distance that the robot may step with the swing leg to reach the balanced home position, and wherein the outer boundary x_capture2 comprises a maximum horizontal distance that the robot may step with the swing leg to reach the balanced home position.

17. The method of claim 9, further comprising:

adding an experimentally determined error value, Δ

x, to the instantaneous capture point, the error value to account for differences between the robot and the model.

18. A computer-readable storage medium storing computer-executable program instructions for controlling a robot having at least two legs including a swing leg, the robot standing on a surface and subjected to a force, the program instructions when executed cause a processor to perform steps of:

receiving a state vector comprising positions and velocities of joints of the robot;

applying a linear motion model to the state vector that models at least a portion of the robot as an inverted pendulum, the linear motion model when applied to the state vector determining an instantaneous capture point on the surface where the robot will step with the swing leg to reach a balanced home position, the balanced home position being a state in which a center of mass of the robot remains substantially over a center of pressure of the robot such that the robot maintains balance; and

outputting a control signal to an actuator that causes the robot to step with the swing leg to the instantaneous capture point.

19. The computer-readable storage medium of claim 18, wherein the instantaneous capture point is located at a horizontal distance x_capture from the center of mass of the robot, where

20. A computer-readable storage medium storing computer-executable program instructions for controlling a robot having at least two legs including a swing leg, the robot standing on a surface and subjected to a force, the program instructions when executed cause a processor to perform steps of:

receiving, by a controller, a state vector comprising instantaneous positions and velocities of joints of the robot;

applying, by the controller, a linear motion model to the state vector that models a first portion of the robot as an inverted pendulum and models a second portion of the robot as a flywheel, the linear motion model when applied to the state vector determining a capture region on the surface where the robot will step with the swing leg to reach a balanced home position, the balanced home position being a state in which a center of mass of the robot remains substantially over a center of pressure of the robot such that the robot maintains balance; and

outputting a control signal to an actuator that causes the robot to step with the swing leg to a point within the capture region.