| dc.description.abstract | This dissertation is the first attempt to demonstrate the use of magnetic-levitation
(maglev) positioners for commercial applications requiring nanopositioning. The key objectives
of this research were to devise the control strategies and motion planning to overcome the
inherent technical challenges of the maglev systems, and test them on the developed maglev
systems to demonstrate their capabilities as the next-generation nanopositioners. Two maglev
positioners based on novel actuation schemes and capable of generating all the six-axis motions
with a single levitated platen were used in this research. These light-weight single-moving
platens have very simple and compact structures, which give them an edge over most of the
prevailing nanopositioning technologies and allow them to be used as a cluster tool for a variety
of applications. The six-axis motion is generated using minimum number of actuators and
sensors. The two positioners operate with a repeatable position resolution of better than 3 nm at
the control bandwidth of 110 Hz. In particular, the Y-stage has extended travel range of 5 mm × 5
mm. They can carry a payload of as much as 0.3 kg and retain the regulated position under
abruptly and continuously varying load conditions. This research comprised analytical design and development, followed by experimental
verification and validation. Preliminary analysis and testing included open-loop stabilization and
rigorous set-point change and load-change testing to demonstrate the precision-positioning and
load-carrying capabilities of the maglev positioners. Decentralized single-input-single-output
(SISO) proportional-integral-derivative (PID) control was designed for this analysis. The effect
of actuator nonlinearities were reduced through actuator characterization and nonlinear feedback
linearization to allow consistent performance over the large travel range. Closed-loop system
identification and order-reduction algorithm were developed in order to analyze and model the
plant behavior accurately, and to reduce the effect of unmodeled plant dynamics and inaccuracies
in the assembly. Coupling among the axes and subsequent undesired motions and crosstalk of
disturbances was reduced by employing multivariable optimal linear-quadratic regulator (LQR).
Finally, application-specific nanoscale path planning strategies and multiscale control were
devised to meet the specified conflicting time-domain performance specifications. All the
developed methodologies and algorithms were implemented, individually as well as collectively,
for experimental verification. Some of these applications included nanoscale lithography,
patterning, fabrication, manipulation, and scanning. With the developed control strategies and
motion planning techniques, the two maglev positioners are ready to be used for the targeted
applications. | en |
