Controller Design for a Parallel Robotic Golf System
Undergraduate Researcher: Nathan Abdalian
Advisor: Dr. Gloria J. Wiens
Sponsor: University of Florida Scholars Program
ABSTRACT
Development
of a two-armed robotic
evaluation system to test golf equipment was explored.
Current robotic testing systems use a single arm approximation for the
golf swing. The single-armed system
is much easier to implement, but does not produce the actual dynamics of a human
golf swing. The two-armed mechanism was developed using a computer
simulation model performed by a previous student.
The model was derived from human biometric data collected using a high
speed camera system and an ADAMS computer simulation. The paper presents the
controller development and design modifications completed on the previous
un-powered/un-controlled prototype.
INTRODUCTION
To
address the shortcomings of current golf test equipment researchers at the
University of Florida have designed a two-armed robotic evaluation system to
test golf equipment (Hunt, Wiens 2003). Current
robotic testing systems use a single armed approximation for the human golf
swing. The main complaint is that
the golf club dynamics of the robot swing do not correlate to the dynamics of
the club during a human swing. The
project uses biometric data for dynamic correlation of a two-armed robot design
(Hunt 2002). Via an ADAMS
simulation model, the kinematics and dynamics of the theoretical design were
validated and tuned. This paper
presents the controller development and design modifications of the first
un-powered/un-controlled prototype (Figure 1).
Figure
1. Two-Armed Simulated Robotic
Mechanism (Hunt and Wiens 2003)
PROTOTYPE
DEVELOPMENT
Increasing the structural stability was the first priority when beginning the golfing robot development. Bachmann, a fellow University Scholar Undergraduate researcher, focused on recreating the prototype using aluminum box tubing to replace the steel components. Analysis was completed on the bearing loads and determined that additional bearings would be implemented to increase the robustness of the design. A one inch steel shaft was used for the main shoulder drive and installed using flanged one inch bearings. The main drive shaft was keyed to allow for connection to the shoulder member. A hardened steel rail and guide block were installed for the right arm mechanism to eliminate the rotation allowed in the previous design. Double bearings were installed at the wrist joint to increase strength and a clamp mechanism was designed to hold the golf club (Bachmann 2004).
During
the mechanical refinement research began using the ADAMS model developed to
determine the torque and speed requirements for the mechanism.
The time from the top of the back swing to impact with the ball was
determined to be .3 seconds. The impact velocity was simulated as being 109mph (current
USGA impact velocity) with the maximum torque required at the shoulder drive
estimated at 110 ft*lbs. Investigation
of methods available to reach this large torque and speed requirements began.
The final design consists of three motors.
The main drive motor simulates the torque developed in the human torso
and lower body and estimated at 110ft*lbs.
The second actuator drives the right elbow joint, which consists of a
linear slide mechanism restricted to 10 inches of travel.
The third actuator rotates the club during the swing, simulating the
human wrist.
Multiple
devices were investigated to develop the high torque and acceleration needed for
the shoulder drive. A Craftsman
Professional 1/2 in. pneumatic impact wrench
with a maximum torque of 450 ft./lbs. forward and 600 ft./lbs.
reverse pneumatic wrench was tested.
The pneumatic wrench was unable to move the shoulder mechanism and thus
eliminated. Pneumatic linear
cylinders were also investigated for the main shoulder drive, but determined to
be limited on the length of travel available.
Next, a cam and bow limb (similar to a compound bow) mounted to a fixed
structure located next to the golfing mechanism was developed.
The bow limb was determined to be limited on the amount of rotation able
to be delivered to the drive mechanism and too dangerous to be used in a final
design. A servo motor was also
researched for the shoulder drive. In
the current phase of the research, the power supply and high torque servo motor
necessary to meet these requirements was not within the project’s budget,
therefore torsion springs were selected for implementing a shoulder driving
force. A torsion spring mount was
created and used to install a 270º torsion spring.
The torsion spring provides additional force as the downswing begins.
The maximum torque provided from the torsion spring occurs upon initial
release and then decreases linearly, which could be used in combination with a
motor to provide additional force while the motor is developing speed.
A mechanical latch was created to allow the user to place the mechanism
at the top of the swing and release the latch, thus initiating the swing.
CONCLUSION
The
two-armed mechanism was refined to prepare for the implementation of actuators.
Actuators were selected for the shoulder drive and the right linear
slide, consisting of a torsion spring and servo motor, respectively.
The servo motor was successively controlled using a computer and LabVIEW®
software.
Future
development consists of mounting the motor to the mechanism via the rack and
pinion track to test the motor performance during the golf swing. A motor can be
used to initiate the latch mechanism developed for the torsion spring shoulder
drive, thus allowing for coordination with the servo motor timing.
A Proportional-Integral-Derivative (PID) algorithm can be implemented for
tracking the robot’s arm motion to follow the corresponding human golfer arm
motion. Finally, the swing can be
recorded using a high speed camera system to validate the results obtained in
the computer simulation model.
ACKNOWLEDGEMENTS
The
author would like to thank Dr. Gloria Wiens for her willingness to support an
undergraduate researcher and guidance during the project.
Special thanks are also extended to the Space Automation and
Manufacturing Mechanisms Laboratory graduate students for the assistance
throughout the project.
REFERENCES
[1]
Bachmann, Jonathan (2004) “Design and Analysis of a Parallel Mechanism
for the Golf Industry: Mechanical
Improvements.”
[2]
Hunt, E.A. (2002) “Design and
Analysis of a Parallel Mechanism for the Golf Industry.”
Technical Report No. SAMM-UF-7.
[3]
Hunt, E.A and Wiens, G.J. (2003) “Design and Analysis of A Parallel Mechanism
For The Golf Industry”, Proceedings of Tenth
International Congress on Sound and Vibration, Stockholm, Sweden, July 7-10,
2003, 8 pp.
[4]
USGA. “USGA To Update Ball Conformance Test.” 19 Dec. 2001 http://www.usga.org/press/2001/2001_85.html