Modeling, Identification and Control of a Pneumatically Actuated, Force Controllable Robot

1
Modeling, Identification and Control of a
Pneumatically Actuated, Force Controllable Robot

• By James E Bobrow Brian W McDonell
• Critiqued by Alan Jennings

2
The System Pneumatic Actuated Robot
Develop mathematical model Describes torque and
position
Identify System Constants
Demonstrate Results
3
Abstract

• What is to be done?
• Model controlling position and force of pneumatic
  robot
• What is new?
• Use of nonlinear dynamics
• Compressed air valve including losses
• How is this done?
• By theory and experiments
• Why is this being done?
• No need of torque/force sensor

4
Introduction

• The Need
• What has been done
• What this research will do
• Derive Nonlinear model
• Experimentally show common inaccuracy
• Model robot as coupled system
• Develop path and force controller
• Demonstrate system strength and weakness

5
Robot Description

• Good
• Pictures of robot
• Describes max force, outstretched force sensor
  location, position resolution more
• Bad
• No brand names or model numbers
• Describing actuator with out fixed points

6
Pneumatic Actuator Model

• Great Derivation
• Starts from energy balance and Fma
• Ignores Heat transfer based on time constant
• Identifies needed measurements

7
Valve Flow CharacteristicsA.K.A. Finding mass
flow rate

• Evaluates standard equation and compares with
  actual test
• Invents test to easily measure by setting volume
  constant
• Poor fit, due to no loss assumption
• Makes new equation that fits
• Does not compare equation with physics
• Two misprints, equation and (t) not in list

8
Control Law for Pneumatic System

• Does not account for changing inertia
• Notation not included in chart
• Describes two methods but uses a third for
  experiments
• Control is not related to robot
• Units switch

9
Experiments

• Results are very clear
• Makes best comparisons possible
• Non-filtered results
• Some things unclear
• how joint torques are measured
• location of movement
• speed

10
Conclusion

• Pulled paper together very well
• Reviewed the steps
• Stressed important areas
• Adds new information
• Claims flow separation is cause of low bandwidth

11
Overall

• Questions?