METHOD FOR GAIT GENERATION AND TRACKING CONTROL FOR A BIPEDAL WALKING ROBOT

• US 20150120044A1
• Filed: 10/31/2013
• Published: 04/30/2015
• Est. Priority Date: 10/31/2013
• Status: Active Grant

• Current Assignee: Disney Enterprises Incorporated (The Walt Disney Company)
• Original Assignee: Disney Enterprises Incorporated (The Walt Disney Company)
• Inventors: CORY, RICK E.
• Granted Patent: US 9,044,859 B2
• Time in Patent Office: Days
• US Class Current: 700/250
• 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
• CPC Class Codes: G05B 2219/40264 Human like, type robot arm
• CPC Class Codes: Y10S 901/01 Mobile robot

Abstract

A control method that executes a numerical non-linear optimization procedure with forward simulation of the full dynamics of the bipedal robot (including ground collisions) to compute a foothold, which produces the desired center of mass velocity during the next step of the walking cycle. The controller includes a gait generator that outputs desired joint trajectories designed to track a desired foothold provided by the foot placement algorithm. A torque calculator is included to output desired joint torques designed to track the desired joint trajectories provided by the gait generator. Additionally, the controller includes an actuator controller that produces the desired joint torques determined by the torque calculator at each joint on the physical robot, which may be a legged robot with torque-controlled joints. The foot placement algorithm may use nonlinear optimization that is solvable using a mathematical model of the dynamics of the controlled robot.

Claims

1. A method for controlling a legged robot with torque-controlled or position controlled joints, comprising:

determining a foot hold location on a support surface for a foot of the legged robot;

generating a set of joint trajectories for joints of the legged robot to achieve the foot hold location;

computing desired torques to provide the set of joint trajectories; and

generating actuator commands to operate actuators on the legged robot to provide the desired torques to move the foot of the legged robot to the foot hold location,

wherein the determining of the foot hold location includes performing nonlinear optimization with forward simulation using the state of the legged robot as a first input and a mathematical model of the legged robot providing constraints for the nonlinear optimization as a second input.

2. The method of claim 1, wherein the determining of the foot hold location is performed based on a predictive forward simulation of full robot dynamics of the legged robot and wherein the generating of the set of joint trajectories is performed as a function of the foot hold location.

3. The method of claim 1, wherein the computing of the desired torques includes partial feedback linearization including applying proportional-derivative control to a subset of the joints.

4. The method of claim 3, wherein the partial feedback linearization includes using a dynamic model of the legged robot to compute partial feedback linearization torques based on desired torques of the subset of the joints that are not dynamically changing in desired trajectories for a number of the joints with dynamically changing trajectories.

5. The method of claim 1, wherein the nonlinear optimization includes performing forward prediction using the mathematical model of the legged robot including robot dynamics and a model of ground collisions between the foot and the support surface including energy losses during the ground collisions.

6. The method of claim 1, wherein the constraints include torque limits of actuators or motors driving the joints of the legged robot.

7. The method of claim 1, wherein the generating of the set of joint trajectories is performed with a stance knee locked in a straight position.

8. The method of claim 1, wherein the generating of the set of joint trajectories involves generating an inner leg angle trajectory using a sine wave parameterized by a phase variable.

9. The method of claim 8, wherein the phase variable is an amount of time elapsing during a step of the legged robot.

10. The method of claim 1, wherein the generating of the set of joint trajectories further comprises defining movement of a swing knee as a Bé

zier Curve parameterized by the inner leg angle trajectory.

11. The method of claim 1, wherein the computing of the desired torques includes using a proportional-derivative (PD) torque controller to track the foot hold location and further using partial-feedback linearization.

12. The method of claim 11, wherein the partial-feedback linearization uses a mathematical model of dynamics of the legged robot to compute the desired torques based on underactuation of the drivers of the joints.

13. A robot controller for providing actuator control signals to motors or actuators used to drive joints or linkages of a legged robot during walking on a surface, comprising:

a foot placement algorithm module generating a foot-hold location on the surface for a foot of the legged robot, wherein the foot-hold location is generated by processing a state of the legged robot and a set of dynamic constraints of the legged robot;

a gait generation algorithm module generating a desired state of the legged robot based on the foot-hold location and the state of the legged robot; and

a torque control module generating commands to the motors or actuators of the legged robot to achieve the desired state of the legged robot including positioning the foot at the foot-hold location on the surface.

14. The robot controller of claim 13, further including a partial feedback linearization module processing the desired state of the legged robot concurrently with a model of the dynamics of the legged robot to produce desired torques for the motors or actuators to achieve the desired state of the legged robot, wherein the torque control module generates the commands based on the desired torques.

15. The robot controller of claim 13, wherein the set of dynamic constraints include torque limits and joint limits.

16. The robot controller of claim 15, wherein the foot-hold location is determined by estimating a 1-step capture point using nonlinear optimization.

17. The robot controller of claim 13, wherein the generating of the desired state of the legged robot is performed by holding the stance knee straight and by defining an inner leg angle to achieve the foot-hold location as a linear Bé

zier Curve.

18. The robot controller of claim 17, wherein the Bé

zier parameterization variable of the linear Bé

zier Curve is defined by a sine wave that is parameterized by a desired swing time and a step time or that is parameterized by stance leg angle.

19. A method for controlling a bipedal robot, comprising:

with a processor running a foot placement algorithm, determining a foot hold location for the biped robot based on a sensed present state of the bipedal robot and on a mathematic model of the biped robot including torque limits and joint limits;

with the processor running a gait generator, generating a desired movement for the bipedal robot to move a foot of the bipedal robot to the foot hold location;

with a torque calculator module run by the processor, determining desired torques for joints of the bipedal robot to achieve the desired movement of the bipedal robot; and

generating actuator commands to operate actuators on the bipedal robot to provide the desired torques on the bipedal robot.

20. The method of claim 19, wherein the determining of the foot hold location is based on a predictive forward simulation of full robot dynamics for the bipedal robot.

21. The method of claim 19, wherein the generating of the movement includes defining joint trajectories for the bipedal robot by generating an inner leg angle trajectory using a sine wave parameterized by a phase variable and by defining movement of a swing knee as a Bé

zier Curve parameterized by the inner leg angle trajectory.

22. The method of claim 19, wherein the determining of the desired torques includes using a proportional-derivative (PD) torque controller to track the foot hold location and further using partial-feedback linearization comprising using a mathematical model of dynamics of the bipedal robot to compute the desired torques.

23. The method of claim 22, wherein the PD torque controller applies control to a subset of the joints of the bipedal robot that are not dynamically changing in desired trajectories and wherein the mathematical model of dynamics is used to compute the desired torques, based on the desired torques for the subset of the joints, for a number of the joints of the bipedal robot with dynamically changing trajectories.