Date of Award

Spring 1-1-2011

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Electrical, Computer & Energy Engineering

First Advisor

Lucy Y. Pao

Second Advisor

Daniel Y. Abramovitch

Third Advisor

Fred Hansen

Abstract

An atomic force microscope (AFM) can provide images with resolution at the atomic scale. The quality and speed of AFM images depend upon the overall dynamics of the AFM system. The behavior of AFMs varies considerably, and currently, the variability causes commercial AFMs to behave unreliably, which slows down and frustrates users. Since the time required to attain a quality AFM image is typically on the order of several minutes or more, substantial motivation exists to reduce the imaging time in AFMs. This thesis discusses combining feedback control with feedforward control for improved speed and performance in tracking applications for AFMs. In particular, two model-inverse-based feedforward architectures are examined. Initial combined feedforward/feedback experimental results show improvement over the feedback-only tracking results, but, being model-inverse-based, these controllers struggle in the presence of system model variation and/or uncertainty. As a result, tracking performance degrades as trajectories vary from the conditions with which the model was identified. To correct for this, a dual-adaptive feedforward algorithm is introduced that adapts on parameters in two feedforward filters. This partnership of adaptive feedforward controllers improved experimental tracking results and robustness to model uncertainties resulting in repeatable high-performance and high-speed tracking throughout the range of the device.

Share

COinS