/
Multi-Policy Control of Biped Walking Multi-Policy Control of Biped Walking

Multi-Policy Control of Biped Walking - PowerPoint Presentation

phoebe-click
phoebe-click . @phoebe-click
Follow
368 views
Uploaded On 2018-02-26

Multi-Policy Control of Biped Walking - PPT Presentation

Eric Whitman 72009 Outline Existing controller Evaluation Resistance to perturbations Speed Control Proposed controller and work towards it 3D Biped Complex System 5 Rigid Links 24 dimensional state space ID: 637166

policy angle swing sagittal angle policy sagittal swing touch control leg touchdown coronal velocity speed angular time plane system

Share:

Link:

Embed:

Download Presentation from below link

Download Presentation The PPT/PDF document "Multi-Policy Control of Biped Walking" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.


Presentation Transcript

Slide1

Multi-Policy Control of Biped Walking

Eric Whitman

7/20/09Slide2

Outline

Existing controller

Evaluation

Resistance to perturbations

Speed Control

Proposed controller and work towards itSlide3

3D Biped (Complex System)

5

Rigid Links

24 dimensional state space

12 control

dimensionsTorque-controlled jointsStatic friction with groundμ=1.0Slide4

Dividing the System

Simplified Coronal Model

Simplified Sagittal ModelSlide5

Sagittal Plane Dynamics

Forward Kinematics

Newton’s Second Law

Non-rotating Torso

For Leg

Substituting:

Simplifying:Slide6

Sagittal Plane Dynamics (Touch Down)

Commanded as an action at any time

Slide7

Sagittal Plane Policy Generation

Ankle Torque Policy

Touchdown PolicySlide8

Full System Sagittal ControlSlide9

Sagittal System Comparison

Major Discrepancies:

Swing Leg Acceleration – Negative initial acceleration

Torso Bob – High late-step accelerationSlide10

Touchdown Prediction

Ankle Torque Policy

Touchdown Policy

Duration of Simulation:Slide11

Time Until Touch DownSlide12

Predicted Angle and Angular Velocity at Touch DownSlide13

Swing Leg Requirements

Must reach touchdown angle at touchdown time

Must touch down without slipping (velocity matching)

Must not accelerate rapidly

Must not touch ground during swingSlide14

Trajectory Generation

Cubic

Spline

Start at current leg angle and angular velocity

Go to predicted touch down angle and angular velocitySlide15

Trajectory Update

Start at currently desired angle and angular velocity

End at new estimate of touch down angle and angular velocity

T

ime

AngleSlide16

Swing Leg Trajectory TrackingSlide17

Swing Knee

Use knee to control swing leg length

Must not touch down during swing

Command foot to 5 cm above ground

Must touch down at appointed time with a straight leg

Command knee angle to straighten linearly:Slide18

Swing Foot Height and Knee AngleSlide19

Coronal Plane Control

Similar dynamics to sagittal plane

Touchdown is different

Switch sides

Pick angle of touch down

Small anglesState not actionSimplified system – counts down and reset to nominal periodFull system – provided by sagittal policySlide20

Full and Simple Coronal ComparisonSlide21

Coronal Swing Leg

Simulate forward to get desired touch down angle

Track with PD servoSlide22

Yaw Control

PD controller on the ankle rotation axis – axis parallel to shin

Does not work well due to large coupling

Must use low gainsSlide23

Perturbation Resistance by TimeSlide24

Failure ModesSlide25

Speed Control by LeaningSlide26

Speed Control By PoliciesSlide27

Parameters

Sagittal Stance

Cost function weights – 2

Nominal Speed

Speed lost at touchdown

Sagittal SwingFoot heightKnee extension rate

Coronal Plane

Cost function weights – 3

Nominal leg angle

Speed lost at touchdown

PD servos

Stance hip – 2Swing hip – 2Swing kneeYaw ankleSlide28

Planned Improvements

Collaborative Coordination

Swing leg policy

More sophisticated sagittal and coronal models

Arms

Better Yaw ControlSlide29

Collaborative Coordination

Each policy supplies V(

t

td

)

Select ttd to minimize sum(V(ttd)) at each time stepEach policy must then know how to act optimally given that

t

td

Slide30

Sagittal Policy with t

tdSlide31

Sagittal Policy with t

tdSlide32

Swing Hip Policy

Dynamic programming

5D state

Angle

Angular velocity

Target AngleTarget Angular VelocityTarget TimeCost Model for TrajectoriesSlide33

Upgraded Simple Models

Unrestrict

the torso angle

Adds 2 more states

Adds an action

Better touch down modelEffect on velocity depends on angleSlide34

Arms

Add an arm swing policy

Probably does not need to participate in collaborative coordination – can be subordinate

Add an estimate of arm swing effect to sagittal policySlide35

Improved Yaw Control

Use adjusted pitch and roll ankle torques to cancel the coupled portion of the rotation torques

Change gain with step

High gain mid-step

Low gain at touchdown

Coordinate with leaning or foot placement?Slide36

Coronal Swing

Can cause trouble if it does not reach its target by touch down

Make a model of tracking error

PD rise time or slew speed

Report expected error to coronal controller

Coronal controller knows cost of this errorSlide37

Robust DP

Normally:

Use:

Or: